UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Bis(ethene)rhodium(I) [beta]-diketonates and related complexes, catalytic and ¹H NMR spectroscopic studies Wickenheiser, Eugene Benedict 1988

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1988_A1 W52.pdf [ 8.3MB ]
Metadata
JSON: 831-1.0059733.json
JSON-LD: 831-1.0059733-ld.json
RDF/XML (Pretty): 831-1.0059733-rdf.xml
RDF/JSON: 831-1.0059733-rdf.json
Turtle: 831-1.0059733-turtle.txt
N-Triples: 831-1.0059733-rdf-ntriples.txt
Original Record: 831-1.0059733-source.json
Full Text
831-1.0059733-fulltext.txt
Citation
831-1.0059733.ris

Full Text

BIS(ETHENE)RHODIUM(I) / ? -DIKETONATES A N D R E L A T E D C O M P L E X E S , C A T A L Y T I C A N D J H N M R S P E C T R O S C O P I C STUDIES By Eugene Benedict Wickenheiser B. Sc. (Chemistry) The University of Regina A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF T H E REQUIREMENTS FOR T H E D E G R E E OF D O C T O R OF PHILOSOPHY in T H E FACULTY OF GRADUATE STUDIES DEPARTMENT OF CHEMISTRY We accept this thesis as conforming to the required standard T H E UNIVERSITY OF BRITISH COLUMBIA August 1988 © Eugene Benedict. Wickenheiser, 1988 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada Date DE-6 (2/88) Abstract A series of bis(ethene)rhodium(I) complexes of /3-diketonates and similar ligands were prepared. The complexes were characterized by elemental analysis, ^B. NMR spec-troscopy and mass spectrometry. Crystal structures are reported for bis(7;2-etheiie)-(l,3(l-ferrocenyl)butanedionato-0.0 ')rhodium(I), 13, (l,5-cyclooctadieiie)(2-acetylphen-oxy-0,0')rliodium(I),18, and some related molecules. 13 18 The complexes are catalyst precursors for the homogeneous hydrogenation of olefins and the hydrosilylation of ketones. The generated hydrogenation catalysts are effective for the reduction of sterically unhindered carbon-carbon double bonds. These catalysts are active in the presence of alcohol, aromatic and carboxylic acid functional groups on the olefin substrate. The catalysts decompose to give a heterogeneously active rhodium precipitate when reducing olefinic substrates which are too sterically hindered. Study of the hydrosilylation reaction revealed that the complexes generate active hydrosilyla-tion catalysts. A series of optically active ligands were tested for their ability to effect asymmetry in the silyl ether products. The bis(ethene)rhodium(I) complexes are fluxional in the N M R time scale due to the motion of the ethylene ligands. A detailed 1 H N M R study was conducted on one of the complexes, bis(7/2-ethene)(2-acetylphenoxy-0,0 ')rhodium(I) 15 to explore the nature of the rearrangement. k A 15 The 1 H NMR study revealed the presence of two different modes of fluxionality. The first type of motion is intramolecular in nature and is due to the rotation of the ethylene ligands about the rhodium-ethylene bond axis. The second type is intermolecular in nature and is due to exchange of the ethylene ligands. This exchange is a measure of the lability of the ethylene ligands. The system allowed the separate study of the ethylene ligands and AG* values were obtained for each ligand for both of the exchange processes. The results of the study indicate the independance of the ethylene ligands with respect to both fluxional processes. iii Table of Contents Abstract ii List of Tables ix List of Figures xi Acknowledgement xiv Abbreviations xv 1 Introduction 1 1.1 Rhodium(I) Catalysis . 1 1.2 Olefin Ligands 4 1.3 The /3-Diketonate Ligand 6 1.4 NMR in Organometallic Chemistry 8 1.5 The Present Study 9 2 Experimental 10 2.1 General Experimental 10 2.2 Characterization 10 2.3 Hydrogenation Experiments . . . 11 2.3.1 Catalyst Poisoning Reactions , 12 2.4 Hydrosilylation Reactions 12 2.4.1 Rhodium Complex Catalysed Hydrosilylation 12 iv 2.4.2 Base-Catalysed Hydrosilylation 15 2.5 Synthesis of Starting Materials and Ligands 15 2.5.1 N,N-Dimethylaminoethylferrocene, FA 15 2.5.2 2-(N,N-Dimethylaminoethylferrocene)carboxyhc acid, FACOOH . 16 2.5.3 2-(N,N-Dimethylaminoethylferrocene)carboxyhc acid methyl es-ter, FACOOMe . 16 2.5.4 2-(N,N-Dimethylaminoethylferrocene)carboxyhc acid ethyl ester, FA-COOEt 17 2.5.5 2-(a-Hydroxyethylferrocene)carboxyhc acid methyl ester, FCMe . 17 2.5.6 2-(a-Hydroxyethylferrocene)carboxylic acid ethyl ester, FCEt . . . 18 2.5.7 Acetylferrocene 18 2.5.8 Diphenylsilane 18 2.5.9 Lithio-acetylferrocene 19 2.5.10 1,3-Dioxobutylferrocene 19 2.5.11 (2-Acetylphenoxy-0,0')diphenylboron, 21 19 2.5.12 3-Benzoyl-(+)-camphor 20 2.5.13 (3-Benzoyl-(+)-camphorato)diphenylboron, 22 20 2.5.14 Pivaloylcamphor 21 2.5.15 Salicyl chloride . 21 2.5.16 (N-Methylbenzyl)salicylamide, 31 21 2.5.17 (2'-Hydroxy)(l-(R)-Q-methylbenzyUmino)ethylbenzene, 32 . . . . 22 2.5.18 1-(1-Phenyl)(l-(R)- Q-methylbenzylamino)buten-3-one, 33 22 2.5.19 Preparation of the Sodium Salts of the Ligands 22 2.6 Preparation of the Rhodium Dimers 23 2.6.1 Bis(ethene)rhodium(I) Chloride Dimer 23 2.6.2 (l,5-Cyclooctadiene)rhodium(I) Chloride Dimer . . . 23 v 2.6.3 (Norbornadiene)rhodium(I) Chloride Dimer 23 2.7 Preparation of the Rhodium Complexes 24 2.7.1 Bis(r/2-ethene)(2,4-pentanedionato-0,0 ')rhodium(I), 1 . . . . . . . 24 2.7.2 Bis(7/2-ethene)(l,3-(l-ferrocenyl)butanedionato-0,0 ')rhodium(I), 13 24 2.7.3 Bis(7?2-ethene)(l,3-(l-phenyl)butanedionato-0,0 ')rhodium(I), 14 . 25 2.7.4 Bis(//-t-butylacetato)bis(norbornadiene)dirhodium 25 2.7.5 Bis(772-ethene)(2-acetylphenoxy-0,0 ')rhodium(I), 15 25 2.7.6 Bis(772-ethene)(salicylato-0)0')rhodium(I)-l/2H20 26 2.7.7 Bis(772-ethene)(sahcylato-0,0 ')rhodium(I), 16 26 2.7.8 Bis(T/2-ethene)(3-benzoyl-(+)-camphorato-0,0 ')rhodium(I), 17 . . 27 2.7.9 (l,5-Cyclooctadiene)(2-acetylphenoxy-0,0 ')rhodium(I), 18 . . . . 27 2.7.10 (l,5-Cyclooctadiene)(l,3-(l,3-diphenyl)propanedionato-0,0 ')rhod-ium(I), 19 27 2.8 NMR Experimental 28 2.8.1 Longitudinal Relaxation Time 7j 28 2.8.2 Total Relaxation Parameter R-y 28 2.8.3 Cross Relaxation Parameter cr 29 2.9 DNMR3 Curve Fitting 29 3 Hydrogenation Studies 31 3.1 Introduction 31 3.2 Preparation of the Ligands 34 3.3 Preparation and Characterization of the Rhodium Complexes 38 3.3.1 Preparation of the Rhodium Complexes using the /3-diketone in Basic Solution 38 3.3.2 Preparation using the Sodium Salt of the Ligands 41 vi 3.3.3 Characterization of the Rhodium Complexes 42 3.4 Hydrogenation Studies 55 3.5 Mechanistic Implications 79 4 Hydrosilylation Studies 86 4.1 Introduction 86 4.2 Results and Discussion 94 4.2.1 Preparation and Characterization of the Optically Active Ligands 96 4.2.2 Asymmetric Hydrosilylation Studies 104 4.2.3 Asymmetric Ligands Based on a-N,N-Dimethylaminoethylferrocene 107 5 N M R Spectroscopic Studies 116 5.1 Introduction 116 5.2 Results and Discussion 127 5.2.1 Ethylene Rotation 130 5.2.2 Dynamic NMR Computer Fitting 143 5.2.3 Error Calculations 152 5.2.4 Ethylene Exchange 153 5.3 Complex 15 in Methanol 166 6 Summary 169 Bibliography 171 A Bis(/i-t-butylacetato)bis(norbornadiene)dirhodium 185 B Diphenylboron Complexes of Ligands 9 and 10 191 B.l (2'-acetylphenoxy-0,0')diphenylboron, 21 191 vii B.2 (3-Benzoyl-(+)-camphorato-0,0 ')diphenylboron, 22 192 C F A C O O H and Related Esters 198 D Base-Catalysed Hydrosilylation 208 D . l Proposed Mechanism 209 vi i i List of Tables 2.1 Coupling constants employed for the DNMR3 fit 30 3.2 Elemental analyses data for 1,3-dioxobutylferrocene and 3-benzoylcamphor 36 3.3 Elemental analyses data for the rhodium(I) complexes 47 3.4 Mass spectrometry and 1 H NMR data for 13 48 3.5 Mass Spectrometry and 1 H NMR data for 14 49 3.6 Mass spectrometry and 1 H NMR data for 15 50 3.7 Mass spectrometry and 1 H NMR data for 16 51 3.8 Mass spectrometry and a H NMR data for 17 52 3.9 Mass spectrometry and 1 H NMR data for 18 53 3.10 Mass spectrometry and 1E NMR data for 19 54 3.11 Effects of steric hindrance on catalysis 61 3.12 Hydrogenation rates 67 3.13 Literature hydrogenation rates 68 3.14 Literature crystal structure data 70 4.15 Hydrosilylation of acetophenone using rhodium/phosphorus-Hgand systems 90 4.16 Hydrosilylation of acetophenone using rhodium/nitrogen-based ligand sys-tems. 92 4.17 Hydrosilylation of acetophenone using the bis(ethene)rhodium(I) complexes 95 4.18 Elemental analyses of the chiral ligands 100 4.19 Mass spectrometry and X H NMR data for 31 101 4.20 Mass spectrometry and J H NMR data for 32 . . 102 i x 4.21 Mass spectrometry and a H NMR data for 33 103 4.22 Hydrosilylation of acetophenone using oxygen/nitrogen-donor ligands . . 105 4.23 Elemental analyses data for FACOOH and its esters 112 4.24 Mass spectrometry and XH NMR data for FACOOH 113 4.25 Mass spectrometry and *H NMR data for FACOOMe 114 4.26 Mass spectrometry and XH NMR data for FACOOEt 115 5.27 R and A values for the protons of ethylene I 140 5.28 R and A values for the protons of ethylene II 143 5.29 Calculated data for ethylene rotation 144 5.30 Data from the DNMR3 fits for ethylene rotation 150 5.31 Activation parameters calculated for ethylene rotation 151 5.32 Observed Data for Dissociation 164 5.33 Calculated data for ethylene dissociation 165 5.34 Dissociation activation parameters 165 A. 35 Mass spectrometry and a H NMR data for bis(^ -t-butylacetato)bis(nor-bornadiene)dirhodium 188 B. 36 Elemental analysis data for the boron complexes 196 B. 37 : H NMR data for the boron complexes 197 C. 38 Elemental analyses data for FCMe and FCEt 205 C.39 Mass spectrometry and J H NMR data for FCMe 206 C. 40 Mass spectrometry and J H NMR data for FCEt 207 D. 41 Base-Catalysed Hydrosilylation Results 210 D.42 Base-catalysed hydrosilylation of 2-cyclohexen-1-one 211 x List of Figures 1.1 The Dewar-Chatt-Duncamson model of metal-olefin interaction 5 1.2 Observed binding modes of/3-diketonate ligands . 7 2.3 Hydrogen Gas Uptake Aipparatus 13 2.4 Modified precatalyst-bnifet dropping device 14 2.5 Diagram of ethylene coupling interactions 30 3.6 Chelation mode of SchifF-base complexes 33 3.7 Ligands employed for preparation of the rhodium(I) complexes 35 3.8 Reaction of Hthio-acetylferrocene and ethyl acetate 37 3.9 Complexes studied as hydrogenation catalyst precursors 40 3.10 Linalool and Geraniol 57 3.11 Hydrogenation of Jinalool 59 3.12 Hydrogenation of geraniol . . . 60 3.13 Catalyst poisoning reaction 64 3.14 Addition of AACA during the reduction of tiglic acid 65 3.15 Stereoview of bis(772-ethene)(l,3-(l-ferrocenyl)butanedionato-0,0 ')rhodium(I), 13, and bond lengths f o r the ligand backbone 72 3.16 Stereoview of (l,5-cyclooctadiene)(2'-acetylphenoxy-0,0')rhodium(I), 18, and bond lengths i o T -the 2'-acetylphenoxide backbone. 74 3.17 Stereoview of (2-acetylphenoxy-0,0')diphenylboron, 21, and bond lengths for the 2'-acetylphencoride backbone . . 75 3.18- Bond lengths in the salicylato ligand of (salicylato-0,0 ')diphenylboron. . 76 xi 3.19 Bond lengths and crystal structure of 2'-hydroxyacetophenone 77 3.20 Possible mechanism for the hydrogenation cycle 81 3.21 *H NMR of bis(772-ethene)(l,3-(l-phenyl)butanedionato-0,0 ')rhodium(I) in methanol-d4 82 3.22 Hydrogenation in toluene 83 3.23 Catalytic hydrogenation of ethylene 84 4.24 Proposed mechanism for the catalytic hydrosilylation of ketones 89 4.25 Crystal structure of N-methylbenzylsalicylamide, 31 106 4.26 Production of a chiral plane in FA derivatives. 108 4.27 Target ligand based on FA 108 4.28 Synthetic route for the FA based diketone ligand 109 5.29 Structure of bis(?/2-ethene)M complexes 120 5.30 Proposed Berry pseudorotation for 4-coordinate complexes 123 5.31 Berry pseudorotation for 5-coordinate rhodium complexes 125 5.32 Ethlene 'inner' and 'outer' protons 128 5.33 Assigned *H NMR spectrum of 15 129 5.34 Assignment of the ethylene protons 130 5.35 The pulse sequence employed for the measurement of Ti . 132 5.36 The temperature dependence of TV for A and C ethylene protons 133 5.37 The temperature dependence of T\ for B and D ethylene protons. . . . . 134 5.38 Ethylene proton T\ experiment 135 5.39 The pulse sequence employed for the determination of R 136 5.40 Experiment 1, C{A}at increasing values of D2 137 5.41 ln(7 - Jo,,) of C{A}versus time (seconds) at -38.5°C 138 5.42 The pulse sequence employed for the determination of er 139 xii 5.43 \n(kACIT) versus 1/T 141 5.44 ki(kBD/T) versus 1/T. 142 5.45 Experimental and calculated spectra 146 5.46 Experimental and calculated spectra continued 147 5.47 ]n(kAC/T) versus 1/T from the DNMR3 fits 148 5.48 \n(kBD/T) versus 1/T from the DNMR3 fits 149 5.49 Spin saturation transfer due to ethylene exchange 154 5.50 Variable temperature 1 H NMR showing intermolecular saturation transfer 155 5.51 The 1 H NMR spectrum of 18 at room temperature 157 5.52 The variable temperature J H NMR study of 18 158 5.53 Proposed reaction mechanism for intramolecular ethylene exchange. . . . 161 5.54 Ethylene I exchange 162 5.55 Ethylene II exchange 163 5.56 Ethylene region of 15 in methanol-^. 167 A.57 Target rhodium(I) complex of 3-pivaloyl-(-f-)-camphor 185 A. 58 Crystal structure of bis(/i-t-butylacetato)bis(norbornadiene)dirhodium. 190 B. 59 Crystal structure of (3-benzoyl-(+)-camphorato-0,0 ')diphenylboron (A). 194 B. 60 Crystal structure of (3-benzoyl-(+)-camphorato-0,0 ')diphenylboron (B). 195 C. 61 Variable temperature a H NMR of the N-methyl region of FACOOH. . . . 199 C.62 The *H NMR of FCMe 203 C. 63 Crystal structure of 2-(a-hydroxyethylferrocene)carboxylic acid methyl ester.204 D. 64 Addition of a five-coordinate silyl anion to a ketone 213 D.65 Possible mechanism for base-catalysed hydrosilylation 213 xiii Acknowledgement I would like to take this opportunity to thank those who have aided me during the course of my work and in the preparation of this thesis. I thank Dr. W.R. Cullen for his supervision and patient guidance, and my advisory commitee: Dr. M. Blades, Dr. B.R. James, and Dr. SchefTer for their helpful suggestions. I extend my thanks to Walter Cicha, Tim Haddad, Cashman Hampton, and Basil Nwata for their insights and suggestions regarding this thesis. Other members of the Cullen group who have helped by stimulating discussion and friendly interaction include Matthew Dodd, Dr. Nam Fong Han, Dr. G. Kang, Abiodun Ojo, Deepthi Hettipathirana, and Tu-cai Zheng. The members of the support services including those working in the analytical, electronic, glass blowing, illustration, mechanical, and NMR divisions of the chemistry department deserve special mention for their dedication, expertise and approachability. I express my appreciation to my wife for her support and for typing much of this thesis, and my family for their encouragement and understanding. I am most thankful to the Lord, my God, without whom none of this would have been possible. xiv Abbreviations A A C A : acetamidocinnamic acid acac : 2,4-pentanedionate ion atm : atmosphere(s) bp : boiling point B M P P : benzylmethylphenylphosphine B P P F A : a-[l',2-bis(diphenylphosphino)ferrocenjd]-ethyldimethylamine bzac : l,3-(l-phenyl)butanedionate ion C H I R A P H O S : Ph 2 PCHCH 3 CHCH 3 PPli 2 C O D : 1,5-cyclooctadiene D l : delay period between NMR experiments D2 : evolution period between radio-frequency pulses (r) dbl : doublet dbzm : l,3-(l,3-diphenyl)propanedionate ion D C T : dicyclooctatetraene 8 : chemical shift (ppm) DIOP ; 2,3-0-isopropyhdene-2,3-dihydroxy-l,4-bis(diphenylphosphino)butane xv D I P A M P : ((o-CH3OC6H4)PhPCH2)2 D M A : N,N-dimethylacetamide e.e. : enantiomeric excess E t : ethyl ether : diethyl ether F A : a-N,N-dimethylaminoethylferrocene F A C O O E t : 2-(a-N,N-dimethylaminoethylferrocene)carboxylic acid ethyl ester F A C O O H : 2-(a-N,N-dimethylaminoethylferrocene)carboxyhc acid F A C O O M e : 2-(a-N,N-dimethylaminoethylferrocene)carboxyUc acid methyl ester F C E t : 2-(a-hydroxyethylferrocene)carboxylic acid ethyl ester F C M e : 2-(a-hydroxyethylferrocene)carboxyhc acid methyl ester G L C : gas liquid chromatography h : hour(s) H z : hertz I : intensity i -P r : isopropyl M : moles per liter M C A : methylcinnamic acid xvi Me : methyl min : minutes mol : mole(s) M P F A : a-[l',2-dimethylphosphinoferrocenyl]-ethyldimethylamine M P P P : methylphenyl-ra-propylphosphine m/z : mass/charge N B D : norbornadiene N M R : nuclear magnetic resonance N O E : nuclear Overhauser effect Ph : phenyl Prophos : Ph 2 PCH 2 CHCH 3 PPh 2 s : second(s) . T F B A : benzoyl-1,1,1-trifluoroacetonate ion T H F : tetrahydrofuran T T A : thenoyl-l,l,l-trifluoroacetonate ion X{Y} : observation of the NMR signal of nuclei 'X' during the irradiation of nuclei xvu Chapter 1 Introduction 1.1 Rhodium(I) Catalysis An overwhelming amount of literature exists describing the chemistry of rhodium(I) complexes [1]. A good portion of this literature deals with the catalytic behavior of rhodium complexes and, of interest here, the area of homogeneous catalysis [2, 3]. The discovery of Wilkinson's catalyst(RhCl(PPh 3)3) has been referred to as a historic landmark in organometallic chemistry [4]. This is appropriate since in the years that followed numerous publications appeared employing Wilkinson's catalyst, and a variety of other rhodium complexes, with phosphorus ligand systems, for a number of catalytic reactions. Examples of the reactions catalyzed by these complexes are hydrogenation [5], hydrosilylation [6], and hydroformylation [7]. Although a large amount of research has taken place in the area of catalysis with rhodium complexes of phosphorus ligands [8, 9, 10], relatively few studies have been undertaken using ligands based on other donor atoms. Rhodium and iridium complexes of ligands such as phenanthroline and bipyridine, which chelate using nitrogen donor atoms, have been reported as useful hydrogenation and hydrosilylation catalysts [11, 12]. Even less work has been reported concerning complexes with ligands based on oxygen donor atoms on rhodium(I). No reports have yet been made regarding the use of rhodium(I) complexes of oxygen-based ligands as hydrogenation catalysts in the absence of phosphines. Some precatalysts with both oxygen- and phosphorus-based ligands are 1 Chapter 1. Introduction 2 described in Chapter 3; in these it is apparent that the catalytically active species no longer contains the oxygen-based ligand. The hydrogenation of olefins, represented in equation 1.1, can be accomplished by using a variety of conditions and numerous catalysts exist which effect this reduction reaction under mild conditions (ambient temperature, 1 atm hydrogen). Other hydrogen sources besides hydrogen gas have also been employed for the reduction reaction [13]; examples of these are sodium borohydride [14] and alcohols [15]. Catalysts which employ alcohols as the hydrogen source are convenient because of their relative ease of handling, their utility as reaction solvents, and the high concentration of available hydrogen. On the other hand, the use of catalysts employing molecular hydrogen as the hydrogen source hold the advantage of no additional side products in the reaction mixture. (1.1) An important question to address when studying a catalytic reaction is whether or not the process is homogeneous. Homogeneous systems are those in which a soluble metal complex is responsible for the catalysis. In heterogeneous systems the catalyst is Chapter 1. Introduction 3 insoluble metal, often deposited as a result of the decomposition of the initial complex. The recent development of polymer bound homogeneous catalyst analogues has made the distinction between homogeneous and heterogeneous catalysis less than clear [16]. The intent of the polymer bound metal complexes is to combine the advantages of the homogeneous and heterogeneous systems. The advantages of the heterogeneous systems are the ease of separation of the catalyst mixture and its subsequent reusability. Hetero-geneous catalysts are also less susceptible to oxidation than homogeneous systems which are often based on air sensitive organometallic species. The advantages of the homo-geneous catalytic systems include greater selectivity and reproducibility as well as the possibility of studying the catalytic system using conventional techniques such as NMR spectroscopy. The use of asymmetric ligands, such as CHIRAPHOS [17], PROPHOS [18], DIPAMP [19, 20], and BPPFA [21], on rhodium(I) have been shown to effect enantioselective reduc-tion of prochiral substrates [22, 23]. Chelating asymmetric ligands give the highest optical yields; chiral monodentate ligands are found to be less effective in this respect [24]. An ex-ample of the asymmetric hydrogenation reaction is the reduction of a-benzamidocinnamic acid with the rhodium complex of chiraphos giving nearly quantitative ( 9 9 % ee) enan-tioselectivity (equation 1.2) [17]. (1.2) S.S R Chapter 1. Introduction 4 Ph2P Ph 2 C H I R A P H O S P R O P H O S M e — c / N M e 2 PPh2 PPh2 D I P A M P B P P F A 1.2 Olefin Ligands Among the most commonly employed ligands are those which coordinate to the metal through an olefinic bond. The first example of an organic moiety bound to a metal was reported by Zeise in 1827 [25]. Zeise's anion ([(C2H4)PtCl3]~) contains an 7/2-bound Olefin coordination has since become a topic of interest because of its relation to the catalytic reactions. It is found that coordination to a metal can render an olefin susceptible to external nucleophilic attack or attack from the metal center itself. The olefin coordination step often involves weakening of the carbon-carbon double bond which can be seen in the infrared spectrum as a decrease in the olefin stretching frequencies Theories have been developed to explain the metal-olefin interaction [28] and the bond model developed by Dewar [29] and by Chatt and Duncanson [30] is the most ethylene ligand and is often used as a classic example of this mode of bonding [26]. [27]. Chapter 1. Introduction 5 Figure 1.1: The Dewar-Chatt-Duncanson model of metal-olefin interaction. widely accepted [4, 27, 31]. This model (figure 1.1) describes the metal-olefin interaction as consisting of two parts: donation of electron density from the 7r-orbitals of the olefin to a metal orbital to form a c-bond, and back donation from,a populated metal d-orbital to the 7r*-orbitals of the olefin. The Dewar-Chatt-Duncanson model has been supported as a workable theory by its ability to rationalize a number of-spectroscopic and structural observations [32, 33, 34, 35, 36]. Rhodium(I) complexes of olefins can be easily prepared and are often employed as catalyst precursors [37]. Rhodium olefin dimers used to prepare catalyst precursors also are used as rhodium(I) sources for in situ catalyst generation in hydrogenation and cat-alytic hydrosilylation reactions [38]. The rhodium dimers are prepared by the addi-tion of ethylene, cyclooctadiene, or norbornadiene to a methanol solution of rhodium trichloride (equation 1.3). The bis(ethene)rhodium(I) chloride dimer is the least sta-ble of the dimers due to the lability of the ethylene, the NBD dimer is more stable, Chapter 1. Introduction 6 and the COD dimer is stable in air. Rhodium complexes can be generated by the ad-dition of a ligand to the appropriate (olefin)rhodium chloride dimer and are often iso-lated as salts of perchlorate, tetrafluoroborate, and hexafluorophosphate ions, such as: [Rh{(PhCH2)MePhP}2(NBD)]+C104 [39,40], [Rh{o-CH30C6H4)PhPCH2}2(C0D)]+BF4-[20], and [Rh(C0D)(PPh3)2]+PFe [41]. (1.3) C 0 H 4 RhCI3 MeOH, H 20 1.3 The /3-Diketonate Ligand The first reports of the preparation of complexes employing /3-diketonates as ligands were made over 100 years ago by Combes [42]. Combes reported the preparation of 2,4-pentandionate complexes of a number of metals (e.g. Na, Be, Mg). At present 6-diketonate complexes have been prepared for nearly all of the transition metal elements [43]. Although the study of rhodium(I) complexes of /3-diketones has been limited, some complexes of this type have been reported. A few rhodium /3-diketonate complexes, related to the work described here, will be discussed in Chapter 3. Chapter 1. Introduction 7 Figure 1.2: Observed binding modes of/3-diketonate ligands. Four major bonding modes commonly encountered with /3-diketonate ligands are summarized in figure 1.2. It is possible to form complexes in which the /3-diketonate ligand acts as a bidentate neutral ligand (A), a monodentate uninegative ligand bound with oxygen (B) or carbon ( C ) , or a bidentate uninegative ligand (D) [44]. The ability to react with metals by different modes of bonding is one of the reasons why these ligands are so versatile. Other types of /3-diketonate-metal interactions have been reported and can be viewed as combinations of the four major modes of bonding [43]. Some reports have been made which imply the existence of an 772-bound complex of the trapped enol form of the /3-diketone, however, no crystal structures have substantiated these claims. Chapter 1. Introduction 8 1.4 N M R in Organometallic Chemistry The use of NMR spectroscopy as a tool for structural and dynamic characterization has increased dramatically with the advent of Fourier transform NMR (FTNMR). The advan-tages of FTNMR over the continuous wave instruments include an increase in sensitivity and improved resolution in the NMR spectrum. The increase in sensitivity means that it is now possible to obtain NMR spectra for complexes which are too expensive to prepare in large quantities or are isolated in low yields. The sensitivity increase also allows the possibility of obtaining spectra of nuclei with low natural abundance without the need for preparing isotopically enriched samples. It is also possible to study the kinetics of reactions with less limitations regarding the stoichiometry of the reagents. Detection of species present in a reaction mixture by distinctive chemical shifts is a fortunate ability of NMR spectroscopy. An example of this type is the determination of the presence of metal hydrides which usually exhibit negative chemical shift values. Metal hydrides can be distinguished from complexed dihydrogen by the measurement of the longitudinal relaxation time which is characteristically short for dihydrogen species [45]. Chiral lanthanide shift reagents have also extended the utility of NMR to the deter-mination optical purity of a sample containing a mixture of enantiomers [46]. Addition of the chiral lanthanide reagent to a mixture of optical isomers causes a difference in the chemical shifts associated with the enantiomers allowing measurement of the rela-tive amounts of the isomers by integration. The lanthanide shift reagent technique is especially useful when the optical rotation value for a particular reagent is unknown. Analysis of fiuxional processes by the use of NMR is an important tool available to organometallic chemists as many organometallic complexes exhibit fluxionality in the NMR time scale [32, 47]. Determination of exchange rates using NMR techniques enables the calculation of the activation parameters for the motion, and allows a number of Chapter 1. Introduction 9 conclusions to be drawn regarding the fluxional process (chapter 5). 1.5 The Present Study The present study was initiated with the discovery that bis(7/2-ethene)(l,3-(l-ferrocenyl)-butanedionato-0,0')rhodium(I) effected the catalytic hydrogenation of 1-octene (equa-tion 1.4). The complex was prepared in high yield by the addition of 1,3-dioxobutyl-ferrocene to a solution of the bis(ethene)rhodium(I) chloride dimer in basic conditions and was found to be reasonably stable. The similar acac derivative bis(772-ethene)(2,4-pentanedionato-0,0 ')rhodium(I) decomposes under the same conditions used for the catalytic hydrogenation of 1-octene. Because of this, a number of complexes similar to the ferrocenyl-diketonate complex were prepared, characterized and studied for their ability as hydrogenation catalysts. The study of the catalytic abilities of the modified /3-diketonate rhodium(I) complexes was extended to embrace the catalytic hydrosilylation reaction. Asymmetric hydrosily-lation reactions were also studied using new asymmetric ligands. The bis(ethene)rhodium(I) complexes prepared were found to exhibit fluxionality in the *H NMR spectra. The complexes were studied by variable temperature NMR and the fluxionahty of one of the complexes was examined in greater detail to gain insight into the nature of the motion. (1.4) Ik A 1 atm.Hz. MeOH Chapter 2 Experimental 2.1 General Experimental The preparation, storage and handling of the compounds used in this study, unless oth-erwise stated, were performed under inert atmospheres of either nitrogen or argon using standard Schlenk and vacuum-line techniques. Solvents were refluxed over appropriate drying agents and distilled prior to use. Diethyl ether and hexane were distilled from calcium hydride; tetrahydrofuran, toluene, benzene, and dimethoxyethane were distilled from sodium/benzophenone; methanol was distilled from magnesium. 2.2 Characterization All of the characterization and analyses of the employed compounds were performed at the Universit}- of British Columbia with the exception of the crystal structure of 2'-(l-hydroxyethylferrocene)carboxylic acid methyl ester which is discussed in the appendix. This structure was performed at the crystallography department of the University of Regina, in Regina, Saskatchewan. Diagnostic 1 H NMR spectra were obtained using either a Bruker WP-80, Varian XL-300, or Bruker WH-400 NMR spectrometer operating at 80 MHz, 300 MHz, and 400 MHz respectively. Gas liquid chromatographs were obtained using Hewlett Packard model 5830A and 5890A gas chromatographs. Mass spectrometry data were collected using a Kratos AES M50 mass spectrometer. Optical rotation measurements were obtained 10 Chapter 2. Experimental 11 using a Perkin-Elmer 141 polarimeter at room temperature using the sodium-D line (589 nm). Elemental analyses were performed by Peter Borda of the University of British Columbia. Analytical data for the new complexes and ligands are presented in the chapters describing their use. All J H NMR spectra displayed in the text were collected on the Varian XL-300 spectrometer (300 MHz) unless otherwise specified. 2.3 Hydrogenation Experiments Hydrogenation experiments were performed using the apparatus shown in figure 2.3. The reaction vessel 'a' was loaded with the solvent and the appropriate olefin. A bucket containing the precatalyst was then placed on the hook and suspended above the olefin solution. The reaction vessel was isolated from the rest of the system by closing tap 'A' and the olefin solution was degassed by three freeze-pump-thaw cycles. During the degassing the entire apparatus was evacuated by leaving all of the remaining taps open. After the degassing the apparatus was loaded with one atmosphere of hydrogen and the reaction vessel was placed in a constant temperature bath. The temperature of the olefin solution was allowed to equilibrate for 10 min. One atmosphere of reference pressure was trapped above the oil surface in the 'U' tube by closing tap 'B'; taps ' C and 'D' were also closed. The reaction was then initiated by dropping the bucket of precatalyst into the olefin solution. As the reaction proceeded, the amount of hydrogen in the gas space above the reaction vessel was depleted and the internal pressure decreased. As the internal pressure decreased, the oil surface exposed to the gas space above the reaction vessel rose with respect to the oil surface exposed to the reference pressure. The amount of hydrogen reacted was determined by measuring the distance it was necessary to move the mercury piston 'E' before the oil levels became equal. The movement of the mercury piston was manually controlled by adding pressure to the right side of the Chapter 2. Experimental 12 mercury surface using tap (D). The distance moved by the mercury piston was read as the volume of hydrogen reacted. Reactions in which the complex was prepared in situ were performed using a similar procedure to the one described above except that the salt of the appropriate ligand was added to the olefin solution and the bis(ethene)rhodium(I) chloride dimer was added by the bucket instead of the previously prepared complex. 2.3.1 Catalyst Poisoning Reactions The catalyst poisioning reactions were carried out in the same fashion as the hydrogena-tion reactions except that once the catalytic reaction was clearly underway a reagent was added to poison the catalyst. The poison used was acetamidocinnamic acid (AACA) and it was added in excess over the initial amount of catalyst precursor added to the reaction mixture (AACA : rhodium RS 5 : 1). The poisoning reactions require that the poison is added to the reaction mixture without spoiling the reaction atmosphere by allowing air in or by changing the pressure inside the apparatus. This was accomplished by modifying the bucket dropper into a corkscrew shape. The corkscrew shaped bucket dropper has the capacity to drop two buckets in succession and to store one during the catalysis prior to the poisoning (figure 2.4). Generally the poison was added after approximately half of the substrate had been hydrogenated. 2.4 Hydrosilylation Reactions 2.4.1 R h o d i u m Complex Catalysed Hydrosilylation The hydrosilylation reactions using the rhodium(I) complexes were carried out using the following protocol. The appropriate ketone was mixed with diphenylsilane and the mixture was degassed using three freeze-pump-thaw cycles. The appropriate rhodium Figure 2.3: Hydrogen Gas Uptake Apparatus. Chapter 2. Experimental 14 Figure 2.4: Modified precatalyst-bucket dropping device. complex was then added with a flow of argon to maintain the inert atmosphere. The mixture was stirred for the reported amount of time and samples were taken from the reaction mixture and hydrolyzed by stirring with 15% aqueous hydrochloric acid. The hydrolysis mixture was stirred for 4 h before analysis. The organic compounds were then extracted with ether, the extracts were combined, dried over magnesium sulfate, filtered, and vacuum distilled. Yields were calculated by GLC and compounds were identified by comparison with authentic samples of the expected products. Optical yields were calculated after measurement of the optical rotation using equation 2.5. («)£ = (2-5) (a)i) = The specific rotation at temperature T measured using the sodium-D line (589 nm). Chapter 2. Experimental 15 ctobs = The observed rotation. I = The cell path length (dm). c = The solution concentration (g/lOOml). Hydrosilylation experiments which involved the in situ generation of the precatalyst complex were performed by adding the bis(ethene)rhodium(I) chloride dimer to a de-gassed solution of ligand (rhodium : ligand ~ 1 : 5), ketone and diphenylsilane. The progress of the reaction was again monitored by GLC of a hydrolyzed sample of the reaction mixture. 2.4.2 Base-Catalysed Hydrosilylation The base catalysed hydrosilylations were carried out by adding the given base to a de-gassed solution of ketone and silane, neat or in the reported solvent. These reactions were performed at 0°C unless otherwise stated. A typical experiment of this type is the hydrosilylation of acetophenone. Acetophenone (2.0 ml, 17 mmol) and diphenylsilane (3.4 ml, 18 mmol) were combined, degassed by three freeze-pump-thaw cycles and placed under an argon atmosphere. Sodium hydride (5 mg of an 80% dispersion in oil, 0.2 mmol) was added and the mixture was stirred for 20 h at room temperature. After hydroly-sis with 10 ml of 10% aqueous potassium carbonate, the product was isolated from the mixture by extraction with diethyl ether and was purified by vacuum distillation. 2.5 Synthesis of Starting Materials and Ligands 2.5.1 N,N-Dimethylaminoethylferrocene, FA The FA was prepared by literature methods and is described elsewhere [48]. Chapter 2. Experimental 16 2.5.2 2-(N,N-Dimethylaminoethylferrocene)carboxylic acid, F A C O O H Butyllithium (12.1 ml of 1.6 M solution in hexanes, 19.4 mmol) was added to a solution of FA (5.00 g, 19.4 mmol) in 100 ml diethyl ether. The solution was stirred overnight, and then cooled to —78°C. Solid carbon dioxide (5 g, 100 mmol) was then added in small portions affording a light yellow precipitate. The mixture was stirred at — 78°C for 2 h and allowed to warm to room temperature. The mixture was filtered under a nitrogen atmosphere and the precipitate was washed with several portions of ether. The precipitate was then converted to the acid and purified by dissolving it in methanol and eluting it on a silica column. The product band was collected and the solvent methanol removed by reduced pressure. The product was dissolved in methylene chloride and dried over magnesium sulfate. The solution was filtered and the solvent removed at reduced pressure to give 5.15 g of the hemihydrate (table 4.23, page 112) as a yellow glassy solid (88% yield). 2.5.3 2-(N,N-Dimethylaminoethylferrocene)carboxylic acid methyl ester, F A C O O M e This was prepared in a manner similar to that described by Grundy et al [49] for the esterification of hindered carboxylic acids. A solution of FACOOH (0.38 g, 1.3 mmol) and potassium carbonate (0.19 g, 1.4 mmol) in 10 ml of dry acetone was prepared and stirred for 30 min. Dimethyl sulfate (0.159 g, 1.26 mmol) was then added and the mixture was refluxed for 3 h. The solution was filtered, concentrated to 1 ml by reduced pressure, and chromatographed on a 7" silica column. A mixture of acetone, ethyl acetate, and triethylamine (75 : 175 : 6) were used as the eluant and gave adequate separation. The major band was collected and the solvent removed at reduced pressure giving 0.195 g (49% yield) of analytically pure product (table 4.23, page 112). Chapter 2. Experimental 17 2.5.4 2-(N,N-Dimethylaminoethylferrocene)carboxylic acid ethyl ester, FA-C O O E t This was also prepared using the method of Grundy et al. A solution of FACOOH (0.96 g , 3.2 mmol) and potassium carbonate (0.48 g, 3.5 mmol) in 45 ml of dry acetone was prepared and stirred for 1 h. Diethyl sulfate (0.49 g, 3.2 mmol) was added and the mixture was refluxed for 4 h. The mixture was filtered and the solvent removed at reduced pressure. The product was dissolved in 2 ml of methylene chloride, loaded onto a 7" silica column and eluted with a mixture of ethyl acetate/triethylamine (97 : 3). The major band was collected and the solvent removed at reduced pressure yielding 0.625 g (60% yield) of the pure ester (table 4.23, page 112). 2.5.5 2-(a-Hydroxyethylferrocene)carboxylic acid methyl ester, F C M e This was prepared using the method of esterification desribed by Alvarez et al [50]; the reaction also effected replacement of the dimethylamino group with a hydroxyl group. A solution of FACOOH (213 mg, 0.708 mmol), sodium bicarbonate (131 mg, 1.56 mmol), and methyl iodide (402 mg, 2.83 mmol) in 1.5 ml of dimethylacetamide was prepared and stirred in the dark for 48 h. The mixture was then hydrolyzed by the addition of 30 ml of 10% aqueous sodium chloride. The product was extracted from the aqueous phase with ether and the combined extracts were dried over magnesium sulfate. The extract solution was filtered and the volume reduced to 2 ml at reduced pressure. The mixture was then chromatographed on silica using petroleum ether and triethylamine (97 : 3) as the eluant. The major band was collected and the solvent removed yielding 88.3 mg (43.3% yield) of the pure ester (table C.38, page 191). Chapter 2. Experimental 18 2.5.6 2-(a-Hydroxyethylferrocene)carboxylic acid ethyl ester, FCEt A solution of FACOOH (0.23 g, 0.76 mmol), sodium bicarbonate (0.14 g, 1.7 mmol), and ethyl iodide (0.48 g, 3.1 mmol) in 1.5 ml of dimethylacetamide were stirred in the dark for 100 h. The reaction was hydrolyzed with 30 ml of 10% aqueous sodium chloride solution. The product was then extracted from the aqueous layer with several portions of ether, the extracts combined and dried over magnesium sulfate. The extract solution was filtered, concentrated to 2 ml at reduced pressure and chromatographed on silica using ether, petroleum ether, and triethylarnine (50 : 50 : 3) as the eluant. The major band was collected and the solvent removed to give 0.13 g (57% yield) of the pure ester (table C.38, page 191). 2.5.7 Acetylferrocene Acetylferrocene was prepared by the Friedel Crafts acylation of-ferrocene and is described elsewhere [51]. 2.5.8 Diphenylsilane This was prepared by the dropwise addition of a solution of dichlorodiphenylsilane (50.0 g. 200 mmol) in 150 ml ether to a suspension of lithium aluminum hydride (5.60 g, 150 mmol) in 250 ml of ether. After the addition was complete, the mixture was refluxed for 4 h. The mixture was hydrolyzed by the addition of 200 ml of 10% HC1 and the product extracted with several portions of diethyl ether. The combined extracts were dried over magnesium sulfate, filtered and vacuum distilled giving the colorless liquid (bp 90° C at 10 mm, literature: 95-97°C at 13mm, 85% yield). Chapter 2. Experimental 19 2.5.9 Lithio-acetylferrocene A bthium diisopropylarnide solution was prepared by the addition of butyllithium (14.4 ml of 1.6 M solution in hexanes, 23 mmol) to a solution of diisopropylamine (2.5 g, 24.7 mmol) in 50 ml of hexane at 0°C. The mixture was stirred for 1 h at 0°C, acetylferrocene (5.00 g, 21.9 mmol) was then added in small portions and the mixture was stirred for an additional hour. The reaction mixture was allowed to warm to room temperature and was stirred overnight. The product was isolated by filtration, washed with several portions of hexane, dried under vacuum and used without further purification. The crude yield was 4.1 g (80%) of the light yellow solid. 2.5.10 1,3-Dioxobutylferrocene Lithio-acetylferrocene (1.5 g, 6.4 mmol) was suspended in 20 ml of ether and cooled to 0°C. Ethyl acetate (0.68 g, 7.7 mmol) was then added dropwise and the mixture was stirred for 4 h at 0°C. The mixture was warmed to room temperature and stirred overnight. The lithium salt of the diketone was isolated by filtration and was washed with several portions of ether. The diketonate salt was converted to the diketone by shaking it with 10% aqueous HC1 solution and the diketone was extracted from the mixture with ether. The combined ether extracts were dried over magnesium sulfate, filtered and the solvent removed at reduced pressure affording a red-orange solid (0.76 g, 44% yield), see table 3.2 (page 36) for elemental analysis data. 2.5.11 (2-Acetylphenoxy-0,0')diphenylboron, 21 Diphenylborinic acid was prepared by the hydrolysis of the (2-aminoethanolato)diphenyl-boron complex (0.50 g, 2.22 mmol) with 35 ml of 15% HC1 and extracted from the aqueous layer with three 20 ml portions of ether. The 2'-hydroxyacetophenone ligand (0.76 g, 5.6 Chapter 2. Experimental 20 mmol) was added to the diphenylborinic acid and the solution was stirred for 2 h. The reaction mixture was then filtered and concentrated by solvent removal. Hexane was added and the solution was cooled to 5°C. Yellow crystals formed (384 mg, 0.88 mmol (40% yield) and were found to be a 1 : 2 mixture of ligand to complex due to the fact that ligand was trapped inside the crystal lattice. Recrystallization from ethanol also afforded crystals whose analysis indicated the presence of 1 molecule of ligand for every 2 molecules of complex (table B.36, page 182). 2.5.12 3-Benzoyl-(+)-camphor A solution of lithium diisopropylamide was prepared by adding butylhthium (22.6 ml of 1.6 M solution in hexane, 36 mmol) to a solution of diisopropylamine (5.1 ml, 36 mmol) in 80 ml of diethyl ether. After 2 h, (+)-camphor (5.0 g, 33 mmol) was added and the mixture was stirred overnight. The solution was cooled to 0°C and ethyl benzoate (6.2 ml, 43 mmol) was added dropwise. After 4 h the ether was removed at reduced pressure and the pinkish solid wras washed with hexane. The solid was hydrolyzed with 25 ml of 20% HC1 and the product extracted with diethyl ether. The ether extracts were dried over magnesium sulfate. The mixture was then filtered and the ether removed at reduced pressure yielding 3.4 g (40%) of pure 3-benzoylcamphor (table 3.2, page 36). 2.5.13 (3-Benzoyl-(+)-camphorato)diphenylboron, 22 The sodium salt of 3-benzoyl-(+)-camphor was prepared by the addition of sodium hy-dride (94 mg, 3.9 mmol) to a solution of 3-benzoyl-(+)-camphor (1.0 g, 3.9 mmol) in 30 ml of anhydrous diethyl ether. The mixture was stirred for 1 h at room temperature. A solution of diphenylborinic acid, prepared by the acid hydrolysis (40 ml 0.1 M HC1) of (2-aminoethanolato)diphenylboron (0.88 g, 3.9 mmol) and extraction with diethyl ether Chapter 2. Experimental 21 (4 x 25 ml), was added dropwise to the solution of the sodium salt of 3-benzoyl-(-(-)-camphor and the mixture was stirred overnight. The solution was filtered, concentrated to 40 ml and refrigerated to afford 1.2 g (73%) of yellow crystals, see table B.36 (page 182) for elemental analysis data. 2.5.14 Pivaloylcamphor Pivaloylcamphor was prepared by literature methods [46]. 2.5.15 Salicyl chloride This was prepared following the literature method [52]. Thionyl chloride (18.0 ml, 247 mmol) in 30 ml benzene was set to reflux and salicylic acid (17.25 g, 125 mmol) was added in 5 g portions every 15 min to give a vigorous reaction with gas evolution. After the addition was complete, the mixture was refluxed for an additional 1 h. The benzene was removed under reduced pressure and the product distilled to give 11.43 g of salicyl chloride (bp 92°C at 15 mm, literature: 92°C at 15 mm, 58% yield). 2.5.16 (N-Methylbenzyl)salicylamide, 31 Methylbenzylamine (2.00 ml, 15.5 mmol) was dissolved in a mixture of triethylamine (20 ml) and ether (20 ml) and cooled to — 78°C. Salicyl chloride (2.24 ml, 18.6 mmol) was added dropwise, stirred for 45 min, allowed to warm to room temperature and stirred for 2 h. The mixture was hydrolyzed with H2O and extracted with ether. The ether extracts were combined, dried over magnesium sulfate, filtered and vacuum distilled (bp 160°C at 1mm), giving 2.18 g (58% yield) of pure product (table 4.18, page 100). Chapter 2. Experimental 22 2.5.17 (2'-Hydroxy)(l-(R)-a-methylbenzylimino)ethylbenzene, 32 The (R)-a-methylbenzylamine (3 ml, 23.3 mmol); 2'-hydroxyacetophenone (3.8 g, 27.9 mmol) and 10 g molecular sieves were combined in 30 ml benzene and refluxed. The reaction was monitored by gas chromatography which indicated the starting amine was spent after 4 h of reaction. The solvent was removed at reduced pressure and the product was distilled (bp 170°C at 1 mm) to give 4.36 g ( 7 9 % yield) of the pure product (table 4.18, page 100). 2.5.18 l-(l-Phenyl)(l-(R)-a-methylbenzylamino)buten-3-one, 33 Benzoylacetone (2.50 g, 15.4 mmol), (R)-a-methylbenzylamine (2.05 g, 17.0 mmol), and 10 g molecular sieves were combined in 50 ml benzene and refluxed. After 6 h, analysis by gas chromatography indicated that the reaction was not progressing any further. The solvent was removed at reduced pressure and the product was vacuum distilled (bpt. 185°C at 1 mm) giving 1.95 g ( 4 8 % yield) of the light yellow oil, see table 4.18 (page 100) for elemental analysis data. 2.5.19 Preparation of the Sodium Salts of the Ligands The sodium salts of the ligands, employed in the in situ hydrogenation reactions and complex syntheses, were prepared by the reacting sodium hydride with the appropriate ligand in diethyl ether. An excess of the ligand was used to ensure complete reaction of the sodium hydride added (ligand : sodium hydride ss 1.5 : 1). An example is the preparation of the sodium salt of 2'-hydroxyacetophenone. The 2'-hydroxyacetophenone (1.00 g, 7.34 mmol) was dissolved in 25 ml of diethyl ether. Sodium hydride (140 mg of an 8 0 % dispersion in oil, 4.9 mmol) was then added in small portions. After the addition was complete the mixture was stirred overnight. The reaction mixture was then filtered Chapter 2. Experimental 23 and the white precipitate was washed with several portions of ether. The ligand salt was dried at reduced pressure and stored under an inert atmosphere. 2.6 Preparation of the Rhodium Dimers 2.6.1 Bis(ethene)rhodium(I) Chloride Dimer Rhodium(III) chloride trihydrate (3.00 g, 11.4 mmol) was dissolved in 30 ml of methanol to which 1.5 ml of water had been added. The solution was then subjected to a slow stream of ethylene gas without stirring. After 16 h the crystals of the dimer were filtered off, washed with cold methanol and stored under an atmosphere of ethylene. The filtrate was treated with ethylene for an additional 16 h and the second crop of dimer crystals treated as the first. 2.6.2 (l,5-Cyclooctadiene)rhodium(I) Chloride Dimer Rhodium(III) chloride trihydrate (2.00 g, 7.60 mmol) was dissolved in 15 ml of 95% ethanol. Freshly distilled 1,5-cyclooctadiene (1.10 g, 10.2 mmol) was then added and the mixture was stirred overnight. The first crop of dimer crystals was filtered off and washed with cold ethanol. The filtrate was treated with 5 ml of isopropyl alcohol and stirred an additional 16 h, the second crop of dimer crystals was washed and combined with the first. 2.6.3 (Norbornadiene)rhodium(I) Chloride Dimer The (norbornadiene)rhodium(I) chloride dimer was prepared in a similar fashion to the COD dimer using rhodium(III) chloride trihydrate (2.00 g, 7.60 mmol) and norbornadiene (1.00 g, 18.9 mmol). Chapter 2. Experimental 24 2.7 Preparation of the Rhodium Complexes The elemental analyses data for the rhodium(I) complexes can be found in table 3.3, page 47. 2.7.1 Bis(7/2-ethene)(2,4-pentanedionato-0,0 ')rhodium(I), 1 A suspension of the bis(ethene)rhodium(I) chloride dimer (50 mg, 129 pmo\) in 15 ml ether was prepared and cooled to — 78°C. Acetylacetone (29 pi, 280 /xmol) was then added followed by the dropwise addition of potassium hydroxide (43 mg, 800 /xmol) in 1 ml of water. The mixture was at stirred at -78°C for 2 h, allowed to warm to room temperature and stirred an additional 2 h. The complex was extracted from the aqueous layer with ether, the extracts combined and the solvent removed at reduced pressure. The complex was recrystallized from ether/hexane giving 55 mg (83% yield) of the pure complex. 2.7.2 Bis(7?2-ethene)(l,3-(l-ferrocenyl)butanedionato-0,0 ')rhodium(I), 13 A suspension of bis(ethene)rhodium(I) chloride dimer (250 mg, 643 /umol) in 15 ml ether was prepared and cooled to — 78°C. The ligand, 1,3-dioxobutylferrocene (382 mg, 141 i^mol), was then added followed by the addition of a solution of potassium hydroxide (180 mg, 3.2 mmol) in 1 ml water. The mixture was then stirred for 4 h, allowed to warm to room temperature and stirred an additional 4 h. The complex was extracted from the aqueous layer with several portions of ether which were combined and concentrated at reduced pressure. Hexane was layered on top of the ether solution to afford 417 mg (87% yield) of the red-orange complex. Chapter 2. Experimental 25 2.7.3 Bis(7? 2-ethene)(l,3-(l-phenyl)butanedionato-0,0 ')rhodium(I), 14 A solution of benzoylacetone (180 mg, 1.11 mmol) and the bis(ethene)rhodium(I) chloride dimer (200 mg, 514 fimol) was cooled to —78°C. A solution of potassium hydroxide (140 mg, 2.5 mmol) in 1 ml water was added dropwise and the mixture was stirred 1 h. The solution was allowed to warm to room temperature and was stirred an additional 1 h. The complex was extracted with several portions of ether, the solvent removed at reduced pressure and the resultant yellow powder recrystallized from hexane to give 293 mg (89% yield) of yellow crystals. 2.7.4 Bis(/x-t-butylacetato)bis(norbornadiene)dirhodium A suspension of the (norbornadiene)rhodium(I) dimer (100 mg, 217 ^mol) and pivaloyl-(-l-)-camphor (110 mg, 493 /xmol) in 20 ml of diethyl ether was prepared, cooled to — 78°C and a solution of potassium hydroxide (120 mg, 2.1 mmol) in 2 ml of water was added dropwise. The mixture was stirred at — 78°C for 2 h, allowed to warm to room temperature and stirred for an additional 1 h. The product was extracted from the aqueous layer with several portions of ether, the extracts combined, and the solvent removed at reduced pressure. The red powder was recrystallized from ether/hexaue to give 167 mg of deep red crystals ( 6 5 % yield) and was identified as the rhodium dimer (page 173, 174). 2.7.5 Bis(77 2-ethene)(2-acetylphenoxy-0,0 ')rhodium(I), 15 Preparation from the Ligand A solution of bis(ethene)rhodium(I) chloride dimer (200 mg, 514 fimol) in 20 ml of hexane was cooled to —78°C and 2'-hydroxyacetophenone (154 mg, 1.13 mmol) was added via syringe. A solution of potassium hydroxide (140 mg, 2.50 mmol) in 1 ml water was then Chapter 2. Experimental 26 added and the solution was stirred for 1 h. The solution was allowed to warm to room temperature and stirred an additional 1 h. The complex was extracted from the aqueous layer, the ether removed and the yellow product recrystalhzed from hexane to give 263 mg ( 8 7 % yield) of yellow crystals. Preparation from the Ligand Salt A solution of bis(ethene)rhodium chloride dimer (100 mg, 257 /xmol) in 20 ml of hexane was cooled to —78°C. The sodium salt of 2'-hydroxyacetophenone (90.0 mg, 569 fimol) was then added and the mixture was stirred for 1 h. The solution was allowed to warm to room temperature and was stirred an additional 30 min. The reaction mixture was filtered and the solvent removed to give 147 mg ( 9 7 % yield) of the product as a yellow crystalline powder. 2.7.6 Bis(772-ethene)(salicylato-0,0 ')rhodium(I) .1/2H 2 0 A solution of bis(ethene)rhodium(I) chloride dimer (200 mg, 514 /imol) in 20 ml of hexane was cooled to —78°C. Salicylaldehyde (120 fil, 1.13 mmol) was then added followed by a solution of potassium hydroxide (140 mg, 2.5 mmol) in 1 ml of water. The mixture was stirred for 1 h, allowed to warm to room temperature and stirred for an additional 1 h. The product was extracted from the mixture with ether and recrystalhzed from hexane to give yellow needle-shaped crystals (229 mg, 7 7 % yield) 2.7.7 Bis(772-ethene)(salicylato-0,0')rhodium(I), 16 A suspension of bis(ethene)rhodium(I) chloride dimer (100 mg, 257 fimol) in 20 ml of hexane was cooled to — 78°C and the sodium salt of salicylaldehyde (90 mg, 624 fimol) was added. The mixture was stirred for 1 h, allowed to warm to room temperature and Chapter 2. Experimental 27 stirred an additional 1 h. The solution was filtered and the solvent removed to give 134 mg (93%) of the pure yellow powder. 2.7.8 Bis(7; 2-ethene)(3-benzoyl-(-f-)-camphorato-0,0 ')rhodium(I), 17 A solution of benzoylcamphor (215 mg, 839 //.mol) and bis(ethene)rhodium(I) chloride dimer (150 mg, 386 //mol) was prepared and cooled to — 78°C. A solution of potassium hydroxide (140 mg, 2.00 mmol) in 1 ml water was then added dropwise and the mixture was stirred for 2 h. The solution was allowed to warm to room temperature and stirred an additional 1 h. The rhodium complex was extracted from the aqueous layer with ether, the solvent removed and the product recrystallized from hexane, giving 262 mg (82% yield) of the orange complex. 2.7.9 (l,5-Cyclooctadiene)(2-acetylphenoxy-0,0 ')rhodium(I), 18 The (l,5-cyclooctadiene)rhodium(I) chloride dimer (225 mg, 456 //mol) was suspended in 20 ml of hexane at room temperature. The sodium salt of 2'-hydroxyacetophenone (200 mg, 1.26 mmol) was added and the solution was stirred for 1 h. The reaction mixture was then filtered through glass wool and the precipitate washed with 25 ml of hexane. The washings and the filtrate were combined and the solvent removed to give 310 mg (98% yield) of pure complex. 2.7.10 (l,5-Cyclooctadiene)(l,3-(l ,3-diphenyl)propanedionato-0,0 ')rhodium(I), 19 The (l,5-cyclooctadiene)rhodium(I) chloride dimer (150 mg, 304 //mol) was suspended in 20 ml of ether and cooled to —78°C. The sodium l,3-diphenyl-l,3-propanedionate (210 mg, 853 //mol) was then added and the mixture was stirred for 1 h. The solution was Chapter 2. Experimental 28 allowed to warm to room temperature and stirred for an additional 1 h. The reaction mixture was filtered through a glass wool filter and the precipitate washed with an additional 10 ml of ether. The washings and the filtrate were combined and the solvent removed at reduced pressure to give 256 mg (97% yield) of pure complex. 2.8 N M R Experimental All specialized NMR experiments were carried out on a Varian XL-300 spectrometer operating at 300 MHz. Routine 1 H NMR experiments were performed using the standard collection and processing parameters supplied by Varian. 2.8.1 Longitudinal Relaxation Time Ta Experiments to determine the longitudinal relaxation time, T\, were performed with the delay time between scans set at a value of approximately 5 2V The delay between 180° and 90° pulses was varied from zero to two times the expected Ta. Values of pulse duration for 180° (92 /isec) and 90° (46 //sec) were obtained from Dr. Orson Chan of this department. Calculation of the values of T\ from the collected data was performed by using the exponential data fit program supplied by Varian using peak intensity as a measure of magnetization. 2.8.2 Total Relaxation Parameter Ri Procedures to determine the direct total relaxation parameter Ri were based on the experiments described by Noggle and Schirmer [53] for use on continuous wave spec-trometers. The delay time between experiments was 5Ti values. The decoupler power was set by experiments in which the decoupler power was increased in a stepwise fashion Chapter 2. Experimental 29 until its increase made no further change in the magnetization of the exchanging nu-clei; the power used for the study was set 0.1 watt higher than the value determined in this way. The delay times between initial saturation and the 90° pulse followed by data collection was varied from zero to the time required for the exchanging system to reach equilibrium. For well resolved peaks the calculation of the values of Ri were made by using intensity as a measure of magnetization. For peaks obscurred by overlap with other resonances integration was used to measure magnetization; the integration boundaries were set carefully, with consideration given to the amount of overlap present. 2.8.3 Cross Relaxation Parameter cr Determination of values of the cross relaxational parameter a were made by using a modified version of the experiment described by Noggle and Schirmer [53]. The same decoupler power employed for the determination of Ri was used. The time employed for irradiation between experiments was also determined from the experiment to determine i?i; this time was set as the time required for the exchanging system to reach equilibrium after the decoupler was tirmed on. The delay time between the equilibrium saturation and the 90° pulse followed by data collection was varied from 0 to 5 T a . 2.9 D N M R 3 Curve Fitting Determination of values of the exchange rate constant for ethylene rotation by curve fitting was achieved by employing the DNMR3 curve fitting program. The coupling constants employed are found in table 2.1 and are labeled in figure 2.5. The trans coupling constants were measured •from spectra acquired at the low exchange limit. Values for cis and geminal coupling constants were used as determined by Cramer [54] for the bis(7/2-ethene)(7/5-cyclopentadiene)rhodium(I) complex because the values were not measureable Chapter 2. Experimental 30 Table 2.1: Coupling Constants Employed for the DNMR3 Fit Employed Parameter Ethylene I Ethylene II Jtrana 13.6 14.0 Jgeminal -0.060 -0.060 8.8 8.8 T2(msec) 48 48 g e m i n a l Figure 2.5: Diagram of ethylene coupling interactions. from the obtained spectra but undoubtedly contributed to line broadening. The low exchange limit, or point at which lowering the temperature further will no longer affect the shape of the peaks, was assigned as having an exchange rate (k) equal to 0. The value of transverse relaxation time was adjusted so that the computer generated spectra at k = 0 would match the low exchange limit spectra. The rest of the computer generated fits were then made by varying k and correcting for the temperature dependence of the chemical shift. The values of k used in the calculation of the activation parameters were chosen from computer generated spectra showing the best agreement with the experimentally observed spectra. Chapter 3 Hydrogenation Studies 3.1 Introduction There have been few reports on the chemistry of bis(ethene)rhodium(I) /3-diketonates since Cramer's initial study on the lability of the olefin ligands of bis(772-ethene)(2,4-pentanedionato-0,0')rhodium(I), 1, [55]. Most of the current research in rhodium(I) chemistry involves phosphorus-based ligands because of the ability of many of these complexes to catalyze processes such as hydrogenation, isomerization, hydrosifylation, and hydroformylation [4]. Rhodium(I) complexes of chelating nitrogen ligands have been used for the catalytic hydrogenation of ketones [56] and olefins [11], and the asymmetric hydrosilylation of ketones [57] (discussed in further detail in chapter 4), however, there are no reports of the use of rhodium(I) /?-diketonate complexes as homogeneous hydro-genation catalysts in the absence of phosphorus-based ligands. Bis(ethene)rhodium(I) /3-diketonate complexes have found use as precursors to a number of complexes. This utility is due to the lability of the ethylene ligands which can easily be displaced by car-bon monoxide [58]. The carbon monoxide ligands can be further displaced by phosphines or phosphites [59]. The ethlyene ligands in some bis(ethene)rhodium(I) diketonate com-plexes can also be displaced to some degree by methanol, an observation which will be discussed in more detail later in this chapter and in Chapter 5. Bis(carbonyl)(2,4-pentanedionato-0,0 ')rhodium(I) 2, which is prepared by the treat-ment of 1 with carbon monoxide, is a catalyst precursor for the hydroformylation of olefins 31 Chapter 3. Hydrogenation Studies 32 I k A Rh oT "To [58, 60]. However, it is probable that the pentanedionato ligand is not present in the cat-alyst, as infrared studies conducted on 2 at conditions milder than those required for cat-alytic hydroformylation indicate only the presence of [Rh5(CO)15]~ and [Rh(CO)4]~ an-ions [61, 62]. Bis(triphenylphosphite)(2,4-pentanedionato-0,0 ')rhodium(I) 3, prepared by the reaction of 2 with triphenyl phosphite [63], is also a hydroformylation catalyst precursor. Again the catalytic species is attributable to one without the pentanedionato ligand; the well known derivative HRh(CO)(P(OPh) 3)3, which is a good hydroformyla-tion catalyst [64], is produced from 3 in the presence of hydrogen, carbon monoxide and free phosphite. Ziolkowski and coworkers also reported the use of complex 3 as a catalyst precursor for arene hydrogenation [65], but again it is likely that the pentanedionato lig-and is lost during catalyst formation. The loss of the pentanedionato ligand was implied from studies made of 3 with D 2 which indicated that the ligand is reductively eliminated under conditions less stringent than those required for catalytic arene hydrogenation [66]. Complexes of Schiff-bases (figure 3.6), similar in structure to /3-diketonate complexes with one of the ketone functionalities being replaced by an imine functionality, are known Chapter 3. Hydrogenation Studies 33 4 Figure 3.6: Chelation mode of Schiff-base complexes. for rhodium, iridium and other metals [67, 68, 69]. The results obtained from the studies of dicarbonyl and 1,5-cyclooctadienyl rhodium(I) complexes of N-alkyl and N-aryl substituted Schiff-bases of general structure 4 (R 1 } R 2 = CO, COD; R 3 = C6H4CH3 — p, C3H7 —n, C 6 H 5 , C 6H 4Cl-p) have not been encouraging [70, 71]. Unsuccessful oxidation reactions with alkyl iodides and incomplete displace-ment of carbon monoxide from the dicarbonyl complexes led the researchers to class the Schiff-base ligands as deactivating [71]. The study of bis(ethene)rhodium(I) derivatives of /3-substituted SchifF-bases of general structure 4 (R 1 } R 2 = C 2 H 4 : R 3 = p-tol, 5, R3 = C H 3 , 6) afforded more interesting results [69]. Complex 5 was found to react reversibly with dioxygen and the olefin ligands of complexes 5 and 6 could easily be displaced by mono- and di-tertiary phosphines and arsines. The bis-phosphine complexes, which have only been prepared via the bis-ethylene complexes (R 1 } R 2 = P(C 6H 5) 3 , R 3 = C H 3 ; Ri, R 2 = P(C 6H 5)2CH 2CH3, R = CH 3), readily undergo oxidative addition reactions with Chapter 3. Hydrogenation Studies 34 methyl iodide. From these studies it is apparent that the complexes with ethylene lig-ands are more reactive than those containing dicarbonyl ligands or the chelating di-olenn cyclooctadiene ligand. The lability of ethylene in complexes such as 1, first noted by Cramer [55], may be a factor in the chemistry of these bis(T/2-ethene)rhodium(I) Schiff-base complexes. It is evident that the lability of the ethylene ligand is also a factor in our study of bis(ethene)rhodium(I) /3-diketonate complexes. 3.2 Preparation of the Ligands The ligands used in these hydrogenation studies are shown in figure 3.7; most are com-mercially available and can be used directly in the preparation of the metal complexes or after minimal purification. Ligands 7 and 11 were prepared and purified by recrystalliza-tion. Ligands 8 and 12 were used as received, and ligands 9, 10 and 2,4-pentanedione were purified by vacuum distillation. The 1,3-dioxobutylferrocene, 7, and benzoylcamphor 11 were prepared by a new method described below. The compounds were identified by mass spectrometry, 1 H NMR and melting point which agreed with literature values [72, 73]. The compounds were isolated as pure samples, giving satisfactory elemental analyses data (table 3.2). Ligand 7 was prepared from 2-hthioacetylferrocene and ethyl acetate. The 2-lithio-acetylferrocene, previously unreported, was prepared by the action of lithium diisopropy-lamide on acetylferrocene (equation 3.6). The lithioacetylferrocene was isolated in high yield (88%) by filtration and purified by washing with hexane. The yellow solid was read-ily soluble in THF and sparingly soluble in diethyl ether. Satisfactory elemental analysis data was not obtained for the hthioacetylferrocene compound, however, its reactivity and products obtained from its use were consistent with the expected formulation. Chapter 3. Hydrogenation Studies 35 Chapter 3. Hydrogenation Studies 36 Table 3.2: Elemental Analyses Data for 1,3-Dioxobutylferrocene 7 and 3-Benzoylcamphor 11 Complex Carbon Hydrogen Oxygen 7 Expected 62.26 5.19 11.86 Found 62.35 5.36 11.90 11 Expected 79.65 7.86 Found 79.90 7.88 (3.6) The reaction of lithio-acetylferrocene and ethyl acetate gave the lithium salt of the diketone as the product. The lithium salt was easily converted to the diketone with aqueous HC1, the highest yield obtained for the diketone being 46.4%. The maximum yield expected from the proposed reaction sequence (figure 3.8) is 50%; this can be explained by the relative acidities of starting material and product. Since the product Chapter 3. Hydrogenation Studies 37 + EtOH Figure 3.8: Reaction of lithio-acetylferrocene and ethyl acetate. alcohol is a stronger acid than acetylferrocene it reacts with the lithioacetylferrocene to give the lithium salt of the alcohol, and acetylferrocene; the acetylferrocene does not react further under these conditions. For ever}' molecule of/3-diketonate produced, a molecule of lithioacetylferrocene is deactivated by protonation resulting in a maximum yield of 50%. This reaction sequence is supported by the yield, by the isolation of the /3-diketone as its lithium salt and by the presence of acetylferrocene in the reaction mixture prior to hydrolj'sis. The benzoylcamphor ligand was prepared in a manner similar to that of 1,3-dioxo-butylferrocene. Camphor was deprotonated by lithium diisopropylamide and then treated with ethyl benzoate (equation 3.7). The reaction sequence is expected to be similar to that proposed for the preparation of 1,3-dioxobutylferrocene, and again the obtained yields did not exceed the predicted 50% maximum. Chapter 3. Hydrogenation Studies 38 (3.7) 1) LiN(i-Pr) 2 3.3 Preparation and Characterization of the Rhodium Complexes Rhodium complexes of the /3-diketone ligands (figure 3.9) were prepared satisfactorily by two methods. The first method made use of the diketone directly in the preparation of the rhodium complex and the second method involved the use of the sodium salt of the ligand. 3.3.1 Preparation of the Rhodium Complexes using the /3-diketone in Basic Solution. These preparations were carried out as described by Cramer for the synthesis of 1 [55]. A solution of the bis(ethene)rhodium(I) chloride dimer and the appropriate ligand in ether was cooled to — 78 °C and an aqueous solution of potassium hydroxide was added (equation 3.8). The reaction mixture was generally stirred for 1-2 h at -78°C and al-lowed to warm to room temperature. Extraction of the product from the aqueous layer Chapter 3. Hydrogenation Studies 39 and removal of the solvent gave the bis(ethene)rhodium(I) diketonate complex which could be further purified by crystallization. Although the complexes were isolated in high yields (%85%), some unreacted dimer or rhodium side product was present in the aqueous phase. This was apparent as the aqueous phase retained color after exhaustive extraction with ether. A precipitate of rhodium metal was also noted in the reaction mixture, indicating that some of the rhodium dimer or product complex had decom-posed. Another disadvantage of this method is that it is not possible to produce the bis(7/2-ethene)(saHcylato-0,0 ')rhodium(I) complex, 16, in its pure form as it bonds wa-ter strongly. The salicylate complex has been isolated as its hemihydrate adduct by this method of preparation. Chapter 3. Hydrogenation Studies 41 3.3.2 Preparation using the Sodium Salt of the Ligands The rhodium(I) complexes of the /3-diketonate ligands can be isolated in near quantitative yields (^97%) in anhydrous conditions by the addition of the sodium salt of the ligand to the bis(772-etliene)rhodium(I) choride dimer or the (l,5-cyclooctadiene)rhodium(I) chlo-ride dimer (equation 3.9). (3.9) R R" i R o - N ° o * a / V R' Chapter 3. Hydrogenation Studies 42 This method of preparation is useful not only because of the high yields and exclusion of water, but also because of the possibility of preparing the complex in situ in the hydrogenation reaction vessel. Purification of the product complex is also unnecessary as filtration of the reaction mixture and removal of the solvent leaves analytically pure complex. 3.3.3 Characterization of the Rhodium Complexes The rhodium(I) complexes were characterized using elemental analysis, mass spectrom-etry, and 1 H NMR spectroscopy. The reported mass spectrometry data does not include peaks of low m/z since they provide little structural information. Unless otherwise stated the 1H NMR data were collected at ambient temperature (18°C). The elemental analysis (table 3.3), mass spectrometry and 1 H NMR (table 3.4) data for complex 13 are consistent with its proposed structure. A singlet at 5.57 ppm with an integral indicative of one proton is assigned to the methine proton 'a'. A doublet of doublets at 4.64 ppm , characteristic for a mono-substituted ferrocene, with an integral indicative of two protons is assigned to the 2' protons of the ferrocenyl ring, b and b'; a similar multiplet at 4.32 ppm, further from the deshielding substituent, is assigned to 'c' and 'c". A broad resonance at 2.90 ppm with an integral indicative of eight protons is assigned to the ethylene protons; this resonance resolves at low temperature. At -43.5°C the resonance at 2.90 ppm gives two multiplets, each with an integral indicative of four protons. Using the rationale of Cramer [55] the multiplet at 3.38 ppm is assigned to the 'outer' protons and the multiplet at 2.36 ppm is assigned to the 'inner' protons. The assignment of 'outer' and 'inner' protons will be discussed in more detail in Chapter 5. A singlet at 2.00 ppm with an integral indicative of three protons is assigned to the methyl T protons. A singlet at 4.10 ppm with an integral of five protons is assigned to the protons of the unsubstituted ferrocene ring, 'd'. Chapter 3. Hydrogenation Studies 43 The mass spectrum of 13 contains a peak indicating the presence of the parent molec-ular ion (428 m/z). Characteristic fragment peaks indicate the loss of one ethylene (400 m/z), two ethylenes (372 m/z), and a peak due to the presence of the free ligand (270 m/z). The elemental analysis (table 3.3), mass spectrometry and X H NMR (table 3.5) data collected for complex 14 are consistent with the expected structure and formulation. A set of multiplets from 7.3-7.8 ppm with an integral indicative of five protons are assigned to the aromatic protons of the phenyl ring. A singlet at 6.02 ppm with an integral indicative of one proton is assigned to the methine proton 'b'. A broad resonance at 3.03 ppm with an integral indicative of eight protons is assigned to the ethylene protons. At -43.5°C the resonance at 3.03 ppm resolves into two multiplets; the multiplet at 3.59 ppm is assigned to the 'outer' protons and the multiplet at 2.54 ppm is assigned to the 'inner' protons. A singlet at 2.14 ppm with an integral indicative of three protons is assigned to the methyl protons, 'd'. The mass spectrum of 14 contains a peak due to the parent molecular ion (320 m/z) and some characteristic fragment peaks. The loss of one ethylene gives a fragment at 292 m/z and the loss of the second ethylene gives the fragment at 264 m/z. A peak due to free ligand is present at 162 m/z. The elemental analysis (table 3.3), mass spectrometry and 1 H NMR (table 3.6) data obtained for complex 15 are consistent with the expected structure and formulation. A set of resonances from 6.95 to 7.15 ppm with an integral indicative of three protons are assigned to the 'a' protons on the benzene ring. A multiplet at 6.31 ppm with an integral indicating one proton is assigned to the 'b' proton; protons ortho to oxygen substituted benzenes resonate in this region of the NMR spectrum [74]. Two broad overlapping resonances at 2.84 ppm and 3.90 ppm are assigned to the ethylene ligands. The low temperature assignment of these protons is discussed in more detail in Chapter 5. Chapter 3. Hydrogenation Studies 44 The mass spectrum of 15 indicates the presence of the parent molecular ion at 294 m/z. Characteristic fragment peaks include ions due to the loss of one ethylene (266 m/z), two ethylenes (238 m/z) and free ligand (136 m/z). The elemental analysis (table 3.3), mass spectrometry and 1 H NMR data (table 3.7) obtained for 16 are consistent with the expected structure and formulation. A singlet at 8.31 ppm with an integral indicative of one proton is assigned to the aldehyde proton 'a'. A set of multiplets between 6.6 and 7.1 ppm with an integral indicative of three protons are assigned to the 'b' protons of the benzene ring. A multiplet at 6.23 ppm with an integral indicative of one proton is assigned to the 'c' proton in the benzene ring. A broad resonance at 3.0 ppm with an integral indicative of eight protons is assigned to the ethylene protons, 'd'. The mass.spectrum of 16 contains a peak corresponding to the parent molecular ion at 280 m/z. Fragment ions are present and correspond to the loss of one ethylene (252 m/z), two ethylenes (224 m/z), and the free ligand (121 m/z). The elemental analysis (table 3.3), mass spectrometry and X H NMR data (table 3.8) are consistent with the expected structure and formulation of 17. A set of multiplets from 7.3 to 7.5 ppm with an integral indicative of five protons is assigned to the aromatic protons on the phenyl substituent. A broad resonance at 2.93 ppm with an integral indicative of eight protons is assigned to the ethylene ligands. At -48.5 °C the ethyletie resonance gives rise to a multiplet at 3.53 ppm assigned to the 'outer' protons and a multiplet at 2.45 ppm assigned to the 'inner' protons. A multiplet at 2.61 ppm with an integral indicative of one proton is assigned to proton 'b' and the multiplets at 2.01 ppm and from 1.0 to 2.0 ppm are assigned to the 'c' protons; these resonances were not resolved well enough to allow separate assignments. Three singlets at 0.90 ppm, 0.80 ppm and 0.72 ppm, each with integrals indicative of three protons, are assigned to the camphor methyl groups 'd', 'e' and T . Chapter 3. Hydrogenation Studies 45 The mass spectrum of 17 contains a peak corresponding to the parent molecular ion at 414 m/z. Characteristic fragment peaks indicate the loss of one ethylene (386 m/z), two ethylenes (358 m/z), and the presence of free ligand (256 m/z). The elemental analysis (table 3.3), mass spectrometry and X H NMR data (table 3.9) obtained for complex 18 are consistent with the structure and formulation expected for this complex. A set of multiplets from 7.0 to 7.2 ppm with an integral indicative of three protons are assigned to the 'a' protons of the benzene ring and the multiplet at 6.29 ppm is assigned to the 'b' proton of the benzene ring similar to the protons of 15. The assignment of the 'c' and 'd' protons is made on the basis that the protons closer to the deshielding ketone functionality should be found further downfield in the NMR spectrum. Using this rationale the multiplets at 4.43 and 4.13 ppm, each with integrals indicative of two protons, are assigned to protons 'c' and 'd', respectively. A singlet at 1.95 ppm with an integral indicative of three protons is assigned to the T protons of the methyl group. Assignment of the resonances at 2.29 and 1.67 ppm in the 1 H NMR spectrum of 18 were made after inspection of the same region of the a H NMR spectrum of 19. Complex 19 is expected to be symmetrical with a C 2 axis passing through the rhodium metal and the methine carbon. The presence of this symmetry element allows the separation of the protons of the cyclooctadiene ligand into three groups: olefinic protons, 'outer' aliphatic protons on the COD ring, and 'inner' aliphatic protons on the COD ring. These groups are in accord with the 1 H NMR spectrum of 19 which contains three separate resonances due to the COD ligand. Had the COD ligand in 19 given rise to two resonances, indicating no difference in chemical shift among the aliphatic protons, the two aliphatic resonances in the NMR spectrum of 18 would have been assigned with regard to the asymmetic nature of the acetylphenoxide ligand, i.e. on the bases of their distance from the deshielding ketone functionality. Since the aliphatic protons (identified by their chemical shift) in Chapter 3. Hydrogenation Studies 46 19 give rise to two separate resonances, the differentiation must be due to the 'inner' or 'outer' orientation of the ring protons. The 'outer' protons are assigned as those giving rise to the low field resonance as this orientation places them closer to the heterocyclic ring and its deshielding anisotropic effect. The orientation of the aliphatic COD protons is illustrated in the crystal structure of 18 which is discussed later in this chapter. The resonances at 2.29 and 1.67 ppm were therefore assigned to proton groups 'e' and 'g', respectively. The mass spectrum of 18 contains a peak corresponding to the parent molecular ion at 346 m/z. Fragment peaks are present due to the loss of the COD ligand (211 m/z) and due to the presence of the free 2'-hydroxyacetophenone ligand (136 m/z). The elemental analysis (table 3.3), mass spectrometry and a H NMR data (table 3.10) obtained for complex 19 are consistent with the expected structure and formulation. A set of multiplets from 7.0 to 7.8 ppm with an integral indicative of ten protons are assigned to the aromatic protons. A singlet at 6.70 ppm with an integral of one proton is assigned to the methine proton 'c'. A broad singlet at 4.42 ppm with an integral indicative of four protons is assigned to the olefinic 'd' protons. The aliphatic COD protons are assigned as described above; the multiplet at 2.31 ppm is assigned to the 'e' protons and the multiplet at 1.68 ppm is assigned to the T protons. The mass spectrum of 19 contains a peak corresponding to the parent molecular ion at 434 m/z. Peaks are present which correspond to the diketone ligand (224 m/z) and the (COD)Rh (211 m/z) fragment ions. Chapter 3. Hydrogenation Studies 47 Table 3.3: Elemental Analyses of the Bis(ethene)rhodium(I) Complexes Complex Carbon Hydrogen Oxygen 13 Expected 50.50 4.94 7.47 Found 50.82 5.03 7.70 14 Expected 52.52 5.35 9.99 Found 52.57 5.48 10.11 15 Expected 49.00 5.14 10.88 Found 48.80 5.15 10.79 16-1/2 H 2 0 Expected 45.70 4.88 Found 45,82, 45.79 4.60, 4.56 16 Expected 47.16 4.68 Found 47.50 4.72 17 Expected 60.87 6.57 7.72 Found 61.10 6.38 7.60 18 Expected 55.51 5.53 9.24 Found 55.64 5.58 9.39 19 Expected 63.60 5.34 7.37 Found 63.54 5.37 7.19 Chapter 3. Hydrogenation Studies Table 3.4: Mass Spectrometry and lE NMR Data for 13 1 m/z I (%) 2 £(ppm) Assignment Protons 3 428 30.9 5.57 a 1 400 10.9 4.64 (multiplet) b,b' 2 372 19.2 4.32 (multiplet) c,c' 2 344 17.1 4.10 d(Cp ring) 5 314 41.1 2.90 e(ethylenes) 8 270* 100.0 2.00 f 3 228 12.1 205 54.7 186 33.6 185 10.6 1The NMR data was obtained at room temperature using CDC13 as solvent. intensity (I) is calculated relative to the base peak (*). 3 The number of protons is calculated by relative integral intensity. Rh Chapter 3. Hydrogenation Studies 49 Table 3.5: Mass Spectrometry and 1E NMR Data, for 14 1 m/z I (%) 2 £(ppm) Assignment Protons 3 320 48.8 7.3-7.8 aromatic 5 292 71.5 6.02 b 1 264* 100.0 3.03 c(eth}'lenes) 8 234 43.4 2.14 d 3 206 65.7 193 16.8 180 19.6 162 12.9 J The NMR data was obtained at room temperature using C D C I 3 as solvent. intensity (I) is calculated relative to the base peak (*). 3 The number of protons is calculated by relative integral intensity. Chapter 3. Hydrogenation Studies 50 Table 3.6: Mass Spectrometry and a H NMR Data for 15 1 m/z I (%) 2 £(ppm) Assignment Protons 3 294 10.88 6.95-7.15 a 3 266 18.94 6.31 b 1 238 22.86 2.84,3.90 c(ethylenes) 8 210 18.80 1.88 d 3 180 16.85 136 58.15 121* 100.0 1The NMR data was obtained at room temperature using toluene-d8 as solvent. intensity (I) is calculated relative to the base peak (*). 3 The number of protons is calculated by relative integral intensity. Chapter 3. Hydrogenation Studies Table 3.7: Mass Spectrometry and lE NMR Data for 16 1 m/z I (%) 2 6 (ppm) Assignment Protons 3 280 30.4 8.31 a 1 252 42.2 6.65-7.1 b 3 224 41.6 6.23 c 1 196 51.2 3.0 d(ethylenes) 8 168 70.9 142 24.8 121 63.9 94* 100.0 1 The NMR data was obtained at room temperature using CDC1 3 as solvent. intensity (I) is calculated relative to the base peak (*). 3 The number of protons is calculated by relative integral intensity. Chapter 3. Hydrogenation Studies Table 3.8: Mass spectrometry and 1E NMR Data, for 17 1 m/z I (%) 2 «5(ppm) Assignment Protons 3 414 15.26 7.3-7.5 aromatic 5 386 30.05 2.93 a(ethylenes) 8 358 23.55 2.61 b 1 328 30.76 2.01 c 2 312 17.34 1.0-2.0 c 2 256 29.57 0.90 d 3 105* 100.0 0.80 e 3 0.72 f 3 a The NMR data was obtained at room temperature using C D C I 3 as solvent. intensity (I) is calculated relative to the base peak (*). 3 The number of protons is calculated by relative integral intensity. Chapter 3. Hydrogenation Studies 53 Table 3.9: Mass Spectrometry and *H NMR Data for 18 1 m/z I (%) 2 <5(ppm) Assignment Protons 3 346 20.9 7.0-7.2 a 3 211 16.7 6.29 b 1 210 16.8 4.43 c 2 208 15.4 4.13 d 2 182 15.4 2.29 e 4 136 49.6 1.95 f 3 121* 100.0 1.67 g 4 1 The NMR data was obtained at room temperature using toluene-d$ as solvent. intensity (I) is calculated relative to the base peak (*). 3 The number of protons is calculated by relative integral intensity. Chapter 3. Hydrogenation Studies 54 Table 3.10: Mass Spectrometry and J H N M R Data for 19 1 m / z I (%) 2 c5(ppm) Assignment Protons 3 434* 100.0 7.84 aromatic 6 286 10.5 7.05-7.20 aromatic 4 285 18.6 6.70 c 1 268 12.5 4.42 d 4 224 31.5 2.31 e 4 223 46.4 1.68 f 4 211 11.8 210 39.8 209 13.8 208 62.1 182 60.6 168 28.4 147 22.3 105 67.8 1 T h e N M R data was obtained at room temperature using toluene-dB as solvent. 2Intensity (I) is calculated relative to the base peak (*). 3 The number of protons is calculated by relative integral intensity. Chapter 3. Hydrogenation Studies 55 3.4 Hydrogenation Studies The complexes employed in this investigation are derived from compounds 7-12. Ligands 7, 8, 11, and 12 are /3-diketones and 2'-hydroxyacetophenone, 9, and salicylaldehyde, 10, can be viewed as the trapped enol form of a /3-diketone. The ligand backbone of the 2'-hydroxyacetophenone and salicylaldehyde chelate complexes are very similar to those of diketonates, an observation supported by crystallographic data to be discussed later. The rhodium complexes of these ligands (7-12) are found to be catalyst precursors for the homogeneous hydrogenation of olefins. These complexes (except 16) are stable under an inert atmosphere and can be stored thus for weeks without noticeable decomposition. Complex 16 decomposes in the solid state during storage under an atmosphere of argon. Chapter 3. Hydrogenation Studies 56 The decomposition product of this complex has not been studied but it is possible that the complex is behaving as the similar Schiff-base complex 5 which reacts with trace amounts of oxygen (discussed in the introduction). For this reason complex 16 was prepared directly prior to its use. Two experimental protocols were followed for the hydrogenation reaction. The first involved adding one of the complexes to a degassed methanol solution of the appropriate olefin in an atmosphere of hydrogen at 30° C. The second involved the in situ generation of the complex. The in situ reactions were performed by adding the bis(ethene)rhodium chloride dimer to a degassed methanol solution of the olefin and the sodium salt of the appropriate ligand. This technique could be used for all of the complexes except 16 as the sodium salt of sahcylaldehyde decomposed in methanol; this was evidenced by the gradual darkening of the solution during storage. A variety of olefinic substrates were studied in an effort to determine the activity and limitations of the catalysts. These studies gave rise to a number of observations which indicated that the catalytic system is both interesting and useful. The catalysts were found to be effective for the reduction of olefins which contain aromatic, alcohol and car-boxylic acid groups. The complexes were also found to effect the selective homogeneous hydrogenation of unhindered carbon-carbon double bonds in the presence of hindered double bonds. The limitations of the catalysts are twofold: sterically hindered double bonds are not reduced homogeneously, and substrates with the ability to chelate the catalyst poison the catalyst and are not reduced. Attempts to reduce a sterically hindered double bond result in decomposition of the catalyst to give a precipitate of rhodium metal which is visibly apparent; the solution quickly changes from a clear yellow color, associated with the complexed rhodium, to a black suspension. This decomposition is consistent with the observation that treatment of the complexes with hydrogen in the absence of Chapter 3. Hydrogenation Studies 57 C H 2 O H Figure 3.10 Structures of linalool (A) and geraniol (B). substrate causes decomposition to rhodium metal. These observations indicate that a coordinating substrate is required to stabilize the catalyst during the catalytic cycle. If the black suspension which formed at the end of the catalytic cycle is allowed to settle, the remaining solution is colorless, indicating that little, if any, complex is remaining. Attempts to detect a complex in the reduction mixture of 1-octene by hydrogen using complex 14 by mass spectrometry were unsuccessful. This analysis was carried out on the residue left after removal of the volatile compounds in the reaction mixture at reduced pressure. The rhodium precipitate produced by complex decomposition heterogeneously hydro-genates the substrate at a rate which is generally much slower than the homogeneously catalysed reductions. Experiments which distinguish between the homogeneous (com-plexed rhodium) and heterogeneous (rhodium precipitate) catalysts are the reduction of linalool and geraniol. Chapter 3. Hydrogenation Studies 58 The reduction of linalool proceeds by the stepwise homogeneous hydrogenation of the sterically unhindered terminal double bond 'a' (figure 3.10) followed by the heteroge-neous hydrogenation of the hindered 'b' double bond. The two processes proceed with different rates, which can be seen in the plot of hydrogen uptake versus time (figure 3.11). Rhodium precipitate was observed to form in the reaction mixture after 423 seconds; this point appears very near to the break in the curve of the uptake plot. The new reaction rate after this break is attributed to the heterogeneous reduction of bond 'b'. The cata-lyst decomposition occurred after 95% hydrogenation of double bond 'a'; hydrogenation continued to complete reduction of bond 'b'. The hydrogenation of geraniol (figure 3.10), which is structurally similar to linalool, but with more steric hindrance, is heterogeneous for both double bonds. This hydrogenation proceeds following an initial formation of rhodium precipitate. A plot of hydrogen uptake versus time for the reduction of geraniol can be seen in figure 3.12. Some substrates illustrating the effect of steric hindrance are listed in table 3.11. Substrates with the ability to chelate the catalyst such as acetamidocinnamic acid (AACA) or maleic acid, are not reduced, nor is a rhodium precipitate formed in the presence of one atmosphere of hydrogen. This situation is unlike that of the sterically hindered substrates where, if they are not reducible homogeneously, the catalyst decom-poses to give the rhodium precipitate, which then reduces the substrate heterogeneously. Furthermore, the addition of AACA to a catalytic hydrogenation reaction stops the re-action, indicating an interaction between AACA and the catalyst which is absent when using sterically hindered substrates. It can be difficult to distinguish between homogeneously and heterogeneously active catalysts (Chapter 1); Crabtree and coworkers have suggested the use of dibenzocyclooc-tatetraene, DCT as a test for catalytic homogeneity [75]. This test works on the basis that DCT has the ability to chelate an active catalyst to form a complex which is stable Chapter 3. Hydrogenation Studies 59 300-1 0 > ! 1 , , , , , 0 1000 2 0 0 0 3 0 0 0 4 0 0 0 5 0 0 0 6 0 0 0 T i m e (s) Figure 3.11: Hydrogen uptake plot for the hydrogenation of linalool catalyzed by complex 14. Chapter 3. Hydrogenation Studies 60 3 0 0 - i 5 0 0 0 Time (s) Figure 3.12: Hydrogen uptake plot for the hydrogenation of geraniol catalyzed by complex 14. Chapter 3. Hydrogenation Studies 61 Table 3.11: The Effects of Steric Hindrance on the Hydrogenation Reaction 1 Substrate Homogeneous Reduction Rhodium Precipitate 0 CH3 -V Or* V 0rv C H 3 COOH V CH3 6 (b)v' CH3 CH3 CH3 V •'Reactions studied at 30°C with the substrate (3.25 x 10 4 moles) in 10 ml methanol catalyzed by complex 14, rhodium/substrate as 1/50, at one atmosphere of hydrogen. Chapter 3. Hydrogenation Studies 62 in the presence of hydrogen effectively sequestering the catalyst from the substrate. A heterogeneous catalyst would not be affected by the presence of the chelating agent and would continue reducing substrate. Acetamidocinnamic acid apparently can act in the same way and can be used as a test for homogeneity on the systems described in this thesis. The effect of the addition of AACA to the ongoing reduction of 1-octene by 14 is shown in the plot of hydrogen uptake versus time (figure 3.13). The addition of the AACA causes the reaction to stop due to poisoning of the catalytic species and is good evidence that the system is homogeneous. The hydrogenation of tiglic acid, 20, using 14 as catalyst, is suspected to be heterogeneous because a rhodium precipitate forms initially, followed by tiglic acid reduction. The addition of AACA to the reduction re-action of 20 by hydrogen lowers the rate but does not stop the reaction which proceeds to completion (figure 3.14). These results support the hypothesis that the 1-octene is being reduced homogeneously and that the tiglic acid is being reduced heterogeneously. The AACA poisoning test was used for a variety of catalyst/substrate systems and in each case the results were consistent with the visual observation, i.e. systems in which a rhodium precipitate was observed in the reaction vessel were shown to be hydrogenating the substrate heterogeneously. Rates for the hydrogenation of a number of substrates using complexes 13-19 are shown in table 3.12. The produced catalysts give varied rates for the hydrogenation of 1-octene. The complexes with bulkier ligands such as 13, with aferrocenyl substituent, and 17, with a benzoyl camphor substituent, give lower hydrogenation rates than 14 which contains a less bulky phenyl group. Complex 16 which employs the salicylaldehyde ligand gives higher hydrogenation rates than complex 15 which has a methyl group in place of the aldehyde-proton of 16. This difference in rates may be due to electronic effects; complex 15 may have a slightly higher electron density in its heterocycle and subsequently on its rhodium atom due to the presence of electron donating methyl group. Chapter 3. Hydrogenation Studies 63 \ D C T The cyclooctadiene complex 18 gives a higher rate of hydrogenation for 1-octene than its bis(ethene) analogue, 15. A possible explanation for this difference is that the COD ligand majr remain bound to the rhodium metal by one of its double bonds during the catalytic cycle unlike in complex 15 where the ethylene ligands can be hydrogenated off or replaced by solvent molecules (discussed later). The presence of a methanol ligand should affect the rhodium metal differently than the presence of an olefin ligand due to the difference in their cr-donor and 7r-acceptor abilities, giving rise to the difference in hydrogenation rates. For comparison, the hydrogenation rates of some of the same substrates, using dif-ferent rhodium catalysts, axe shown in table 3.13. It is possible to draw quantitative conclusions regarding relative rates in three situations: the first is the case of comparing Chapter 3. Hydrogenation Studies Figure 3.13: The in situ poisoning reaction of the catalyst during 1-octene reduction Chapter 3. Hydrogenation Studies H C H 3 W CH3 COOH 20 400 -1 — 300 o •o o 5 200 c bC O L. X 100 k A A C A added I 2000 4000 T i m e (s) 6000 8000 Figure 3.14: Addition of AACA during the reduction of tiglic acid. Chapter 3. Hydrogenation Studies 66 two catalysts under the same reaction conditions (i.e. metal and substrate concentra-tions, temperature, solvent, hydrogen pressure), the second is that of a catalyst system which reduces a particular substrate at a higher rate with lower concentrations of both substrate and catalyst than another system, and the third is the comparison of systems in which the kinetics have been established. Using these criteria it is possible only to make quantitative comparisons for the rate of hydrogenation of cyclohexene. Complex 14 effects the hydrogenation of cyclohexene at a higher rate than RhCl(PPh3)3 and the reaction conditions for the reduction using 14 employ lower concentrations for both cat-alyst and substrate. Complex 14 also effects the hydrogenation of 1-octene at a higher rate, employing lower catalyst and substrate concentrations, than for the hydrogenation of the similar 1-hexene using RhCl(PPh3)3. Qualitative comparisons between the system studied in this thesis and the quoted literature results seem to indicate that for the catalyst/substrate systems described here the rhodium(I) /3-diketonate catalysts have comparable and higher rates dependant on the substrate. Therefore, on the basis of hydrogenation rates, the new system may be classified among these literature catalysts which have been recognized as highly active [99]. It is interesting to note that bis(7/2-ethene)(2,4-pentanedionato-0,0 ')rhodium(I), 1, does not act as a homogeneous hydrogenation catalyst precursor under the same condi-tions as 13-19. When the hydrogenation reaction was attempted (30°C, 1 atmosphere of hydrogen), using 1-octene as substrate, 1 decomposed giving rhodium metal. The fun-damental difference between 1 and the complexes which produce homogeneous catalysts is the presence of an aromatic ring adjacent to, or incorporated in, the heterocyclic ring formed by chelation. Crystal structures of 13 and 18 indicate the presence of conjugation between the phenyl ring and the heterocycle which may affect the ability of the catalyst during the hydrogenation cycle. Complexes 13-20 which have aromatic rings adjacent Chapter 3. Hydrogenation Studies 67 Table 3.12: Maximum Rate of Hydrogenation for the Catalyst Systems.1 Precatalyst [Rhodium] (M) Substrate [Substrate](M) Maximum Rate 2 (h 1) 13 4.80 x 10"4 1-octene 1.98 x 10"2 76.5 14 2.49 x 10"4 1-octene 3.25 x IO"2 1580 14 5.22 x 10"4 fumaric acid 3.10 x IO"2 16.2 14 3.39 x 10"4 styrene 3.23 x IO"2 185 14 3.11 x 10"4 cyclohexene 3.25 x IO"2 1308 14 5.35 x 10~4 tiglic acid 3.41 x I O - 2 267 14 6.89 x 10"4 M C A 3 1.99 x 10~2 50.1 14 ' 5.99 x 10~4 linalool 1.64 x IO"2 'a' 237 V 23.6 14 10.5 x IO"4 geraniol 1.61 x 10~2 235 15 9.18 x 10"4 1-octene 3.25 x IO"2 154 154 8.16 x 10"4 1-octene 3.25 x IO"2 24 16 7.10 x 10~4 1-octene 3.25 x IO"2 391 ' 17 8.69 x 10"4 1-octene 3.25 x 10~2 63.8 18 3.18 x 10"4 1-octene 3.25 x 10~2 612 19 2.07 x IO"4 1-octene 3.25 x IO" 2 277 1 All reactions carried out in 10 ml of methanol and 1 atmosphere of hydrogen at 30°C. 2Calculated from the maximum slope of the hydrogen uptake plot for the reaction by (mol substrate reduced)(mol catalyst) _ 1 h _ 1 . 3 MCA: methylcinnamic acid. 4Reaction carried out in 10 ml of toluene and 1 atmosphere of hydrogen at 30°C. Table 3.13: Hydrogenation using Other Rhodium(I) Catalysts Precatalyst [Rhodium](M) Substrate [Substratej(M) Rate (h-1) 1 Ref. [Rh(COD)(PPh3)2]PF62 0.2 mM 1-hexene 0.5 M 1500 [41] RhCl(PPh3)3 3 1.25 mM 1-heptene 0.313 M 165 [76] [Rh(COD)(PPh3)2]PF6 4 0.2 mM cyclohexene 0.5 M 10 [41] RhCl(PPh3)3 5 0.625 mM cyclohexene 0.625 M 332 [76] RhCl(PPh3)3 6 0.108 mM linalool 6.48 mM 14 [77] Hh 2 Cl 2 (C 8 H 1 4 )4 / 2-alkylaminopyridine 7 0.178 mM cyclohexene 0.45 M 2500 [78] Calculated by (mol substrate reduced)(mol catalyst) -1h_1. 2Reaction at 20°C in 16 ml benzene (0.1 M triethylamine), one atmosphere of hydrogen. 3Reaction at 25°C in 80 ml benzene, 0.65 atmospheres of hydrogen. 4Reaction at 20°C in 16 ml benzene (0.1 M triethylamine), one atmosphere of hydrogen. 5Reaction at 25°C in 80 ml benzene, 0.65 atmospheres of hydrogen. 6Reaction at ambient temperature in 20 ml benzene, one atmosphere of hydrogen. The rate given is for the reduction of double bond 'a' discussed earlier. 7Reaction at 30°C in 22 ml ethanol, one atmosphere of hydrogen, rhodium/ligand = 1/4, ligand = 2-(N-n-butylamino)pyridine. Chapter 3. Hydrogenation Studies 69 to the heterocyclic ring also may be stabilized by conjugation of the two rings. The conjugation effect is evident in the crystal structure of 13, in which the ferrocenyl ring is nearly coplanar with the heterocycle. This near coplanarity has been suggested as an indication of conjugation between the rings by Hon et al for complexes of vanadium [79], copper [80], palladium [81], and zinc [82] with ligand 8; a shortening of the bond between the aromatic ring and the heterocycle gave further evidence of the conjugation. All of the complexes studied by Hon et al revealed that the phenyl ring was nearly coplanar with the heterocycle. A number of complexes of a variety of metals and similar ligands have shown this same effect. Some of the results are summarized in table 3.14. The crystal structure of 13 was performed by Dr. Steve Rettig of the University of British Columbia and is shown in figure 3.15. The crystal structure confirms the expected molecular structure in which the /3-diketonate ligand chelates a square-planar rhodium metal ion also containing two cis-ethylene ligands which stand perpendicular to the plane of rhodium coordination. Possible conjugation between the aromatic ferro-cenyl cyclopentadiene ring and the heterocyclic chelation ring is indicated by their near coplanarity. The angle between these rings is found to be 11.6 degrees. It is not possible to conclude the presence of a shortening effect in the carbon-carbon bond between the cyclopentadiene and heterocyclic rings because of the large standard deviation associ-ated with this bond length. Derealization within the heterocyclic ring is implied by the similarities in bond lengths between rhodium and oxygen (Rh-O(l), Rh-0(2)), carbon and oxygen (C(ll)-O(l), C(13)-0(2)) and the carbon-carbon bonds (C(ll)-C(12), C(12)-C(13)). The lengths of these related bonds are within the experimental error determined for them; the error in the bond lengths is low enough to determine a difference in the order of these bonds. No significant difference is apparent in the rhodium-ethylene bond distances nor in the carbon-carbon bond distances in the ethylene ligands. Significant differences are not present between the bond lengths in complex 13 and Chapter 3. Hydrogenation Studies Table 3.14: Angles Between Aromatic and Heterocyclic Chelate Rings Compound Angle Between Rings (degrees) Reference V(bzac)2 6.5,19 [79] Cu(bzac)2 14.3 [80] Pd(bzac)2 23.2 [81] Zn(bzac)2OCH2CH3 7.5,3.8 [82] Y(bzac)3H20 16.0,21.7,15.6 [83] Ho(dbzm)H20 10.3 [84] Rh(dbzm)(C3H4)4 a [85] Sn(bzac)2 30.4 [86] Ph3Sn(dbzm) 6.1,3.7 [87] Ph3(CO)Rh(TTA) 5.4 .[88] Rli(TFBA)(CO)2 12.9 [89] "Angle not specified but given structure shows near coplanarity. bzac: l,3-(l-phenyl)butanedionate dbzm: l,3-(l,3-diphenyl)propanedionate TTA : thenoyl-l,l,l-trifluoroacetonate TFBA: benzoyl-l,l,l-trifluoroacetonate Chapter 3. Hydrogenation Studies 71 those found in the crystal structure of the acac complex (bis(r;2-ethene)(2,4-pentane-dionato-0,0')-rhodium(l)) [90], however, the differences may not be noticeable as the standard deviation in the bond lengths is high in the crystal structure of the acac complex (estimated standard deviation % 0.02 A). The crystal structure of (l,5-cyclooctadiene)(2'-acetylphenoxy-0,0 ')rhodium(I), 18, was determined by Dr. Rettig, and is shown in figure 3.16. The structure confirms the expected mode of coordination of the ligand which chelates a square-planar rhodium atom. The cyclooctadiene ligand is coordinated to the rhodium with the double bond axes perpendicular to the rhodium-coordination plane as expected (more examples of the structure of rhodium(I) olefin complexes are given in Chapter 5). Observation of the 2'-acetylphenoxide ligand backbone reveals that the aromaticity of the benzene ring has been disrupted significantly, giving rise to two localized double bonds between C(ll) and C(12), and C(13) and C(14). The bonds between C(14)-C(9), C(9)-C(10), C(10)-C(ll) and C(10)-C(15) are between the lengths expected for single (1.541(3) A) and aromatic carbon-carbon bonds (1.395(3) A) [91]. The Rh-O(l) bond is significantly shorter than the Rh-0(2) bond, and.C(15)-0(2) bond is significantly shorter than the C(9)-0(l) bond giving rise to a long-short-long bond length pattern about the heterocycle. A crystal structure of the diphen}dboron analogue of 18 has also been solved by Dr. Rettig and is shown in figure 3.17. Boron is expected to have less 7r-interaction with the chelating ligand than rhodium because of the absence of d-orbitals in boron, however, the ligand backbone reveals similarities between the rhodium and boron complexes. The carbonyl oxygen is bound to the boron atom with a significantly longer bond than the phenoxy-oxygen, similar to the rhodium complex. Again, the carbon-oxygen bonds differ significantly with the carbonyl carbon-oxygen bond being the shorter. A localized double bond is apparent between C(3) and C(4), however, it is only significantly different from the C(2)-C(3) bond; the data does not distinguish between the C(3)-C(4) bond and the Chapter 3. Hydrogenation Studies 72 Figure 3.15: Stereoview of bis(T/2-ethene)(l,3-(l-ferrocenyl)butanedionato-0,0 ')rhod-ium(I), 13, and bond lengths for the ligand backbone. Chapter 3. Hydrogenation Studies 73 C(4)-C(5) bond. A diphenylboron complex of benzoylcamphor (ligand 11), (3-benzoyl-(-l-J-camphorato-OjO ')diphenylboron (22), has also been prepared and studied by x-ray crystallography. The structure of 22 shows that the heterocycle formed by chelation is not planar, indicating little aromatic character in the ring. Preparation and characterization of 21 and 22, and the crystal structure of 22 are contained in the appendix. The crystal structure of the similar (sahcylato-0,0 ')diphenylboron complex [92] is comparable to that of the acetylphenoxide complex (figure 3.18). Again the ligand coor-dinates in a lopsided fashion with significant differences in boron-oxygen bond lengths. The benzene ring contains only one significantly different carbon-carbon bond between C(3) and C(4). This localized bond is significantly different from the bonds C(l)-C(2) and C(2)-C(3). A localized double bond cannot be verified between C(5)-C(6) for either the acetylphenoxide or the salicylato diphenylboron complexes. A fortunate consequence of the solution of the crystal structure of (2'-acetylphenoxy-O,0')diphenylboron was the acquisition of the crystal structure of the free 2'-hydroxyl-acetophenone ligand which cocrystallizes with the complex. The structure of the ligand is shown in figure 3.19. Determination of the crystal structure from a sample of pure ligand would have been less convenient as it has a melting point of +4°C. The structure reveals that the hydroxyl proton occupies the chelation site between the two oxygen atoms. An interesting feature of the structure is a signficantly short bond between C(4s) and C(5s); this bond is distinguishable from the C(3s)-C(4s) bond but not from the C(5s)-C(6s) bond. In the rhodium complexes the analogous bond is significantly longer than the neighbouring carbon-carbon bonds. Other interesting differences between the free and complexed ligand include a significant difference in the C(l)-0(2) bond length between the free ligand and ligand coordinated to rhodium (this bond is shorter in the rhodium complex), and a significant difference in the C(l)-C(2) bond length between the free ligand and the ligand coordinated to rhodium (this bond is longer in the rhodium Chapter 3. Hydrogenation Studies 74 Figure 3.16: Stereoview of (l,5-cyclooctadiene)(2 '-acetylphenoxy-0,0 ')rhodium(I), 18, and bond lengths for the 2'-acetylphenoxide backbone. Chapter 3. Hydrogenation Studies 75 Figure 3.17: Stereoview of (2'-acetylphenoxy-0,0 ')diphenylboron, 21, and bond lengths for the 2'-acetylphenoxide backbone. Chapter 3. Hydrogenation Studies 76 Figure 3.18: Bond lengths in the salicylato ligand of (salicylato-0,0 ')diphenylboron. complex). Both rhodium(I) crystal structures indicate the possibility of an increase in 7r-electron density in the chelate heterocycle. The increase in 7r-electron density can be attributed to the 7r-donor ability of the benzene and cyclopentadiene rings [4, 93]. This effect may be responsible for the catalytic ability of these complexes as compared to the acac rhodium(I) complex which decomposes under hydrogenation conditions (one atmosphere of hydrogen, rhodium : substrate % 1 : 50). A possible 7r-interaction between the ferrocenyl ring and the heterocyclic ring is in-dicated in the structure of 13 by the near coplanarity of the two rings, similar to the previously cited examples. Evidence of a Tr-interaction between the benzene ring in 18 and the heterocycle is shown in the localization of double bonds in the benzene ring and changes in some of the bond distances in the heterocycle. A shorter C(l)-0(2) bond in the rhodium complex than in the free ligand implies more 7r-character in the heterocycle Figure 3.19: Bond lengths and crystal structure of 2'-hydroxyacetophenone. Chapter 3. Hydrogenation Studies 78 and significantly shorter bonds between C(l) and C(2) in the benzene ring of the free ligand implies 7r-electron density donation to the heterocyclic ring. Comparison of the boron complexes to the rhodium complexes reveals that the rhodium complexes have more differences from the free ligand than the boron complexes. These differences can be attributed to the ability of the rhodium metal to interact with the 7r-electron den-sity of the heterocyclic ring with its d-orbitals; these orbitals are missing in the boron complexes. The lack of aromaticity indicated in the structure of 22 (presented in the appendix) can also be attributed to boron's inability to interact with the 7r-system of the diketonate ligand. An increase in the 7r-character of the heterocyclic ring would increase the electron density in the metal ion; such an increase allows greater metal-olefin interaction and is expected to strengthen the metal-olefin bond. An increase in electron density on the metal also increases its ability to undergo oxidative hydrogen addition which is a required step in the catalytic hydrogenation cycle for both neutral [94] and cationic rhodium catalysts [95]. An increase in metal-olefin backbonding also destablizes the carbon-carbon double bond, increasing the possibility of hydrogen insertion. The lack of these effects may explain why the acac rhodium(I) complexes are unable to act as homogeneous hydrogenation catalysts. Aromatic substituted /3-diketonates have been reported to chelate more strongly than those with methyl end-groups by studying formation constants with metals such as Fe, Co, Ni, Pd, Cu [96, 97]. These chelation strengths were attributed to the ability of the phenyl groups to 'shield' the metal ion from nudeophilic attack more effectively than the methyl end groups. This effect may contribute to the stability of the studied complexes. Chapter 3. Hydrogenation Studies 79 3.5 Mechanistic Implications A possible general catalytic reaction scheme is outlined in figure 3.20. As previously mentioned, the bis(ethene)rhodium(I) complexes may be prepared in situ (A) by the addition of the bis(ethene)rhodium(I) chloride dimer to the sodium salt of the ligand or the complex may be used directly (B). The presence of the solvent methanol displaces ethylene from the complex to some extent. Evidence for the ethylene displacement for complex 15 can be seen in the 1 H NMR spectrum in methanol- d4 in which free ethylene is evident at 8 = 5.2 ppm (free ethylene in solution has a chemical shift of 8 = 5.2 ppm [98]) (figure 3.21). The ethylene displacement from complex 15 appears to occur specifically for the ethylene trans to the oxygen bound to the phenyl ring; this is evident in the low temperature 1 H NMR spectrum and will be discussed in more detail in Chapter 5. These complexes are also catalytically active in toluene, e.g. complex 15 effects the catalytic hydrogenation of 1-octene in toluene (figure 3.22). An interesting feature of the reduction in toluene is the rate, which is considerably slower than the rate obtained when the reaction is carried out in methanol (rate in methanol/rate in toluene ~ 6). This difference suggests a role for the solvent in the catalytic cycle. Production of the substrate-coordinated rhodium species may proceed by two routes: ethylene may be displaced by solvent, evident in the NMR experiments, which may be further displaced by substrate, and ethylene may be directly replaced by substrate. The possibility for direct ethylene replacement by substrate is supported by the ability of the catalyst to effect reduction in toluene, a non-coordinating solvent [99]. Further support for the direct replacement of ethylene by substrate can be deduced from the lability of the ethylene ligands; a 1 H NMR study (Chapter 5) establishes the lability of the ethylene ligands of 15 in toluene. Labile ethylene ligands are likely to be replaced by substrate due to mass action which would predict that the abundance of substrate favours the formation Chapter 3. Hydrogenation Studies 80 of the substrate-coordinated species under the conditions used for the catalytic runs. Reduction of the ethylene ligands is also a viable route to production of the catalyst. Complex 15 effects the hydrogenation of ethylene when exposed to a mixture of ethylene and hydrogen (1 : 1) at one atmoshere of pressure. It should be noted, however, that this reaction proceeds homogeneously for 12 turnovers only, before catalyst decomposition occurs and rhodium precipitates from the reaction mixture. After the precipitation of rhodium, ethylene reduction proceeds at a slower rate. The two-rate effect is shown in the plot of hydrogen/ethylene reacted versus time in figure 3.23. This reaction is interesting since Wilkinson's catalyst does not hydrogenate ethylene [100]. No reaction was observed when the ethylene reduction was attempted under the same conditions using toluene as the solvent. This is another indication that the solvent has a role in the catalytic process. A consistent trend in the cycle is the decomposition of the catalyst to rhodium metal at the end of the reaction as the concentration of substrate becomes depleted. The appearance of the rhodium precipitate is generally evident after 90% of the substrate has been hydrogenated. The hydrogenation reactions continue, if allowed, to complete conversion of the substrate to the corresponding alkane. Again, it seems that coordinating substrate is required to stabilize the catalyst since catalyst decomposition occurs when the substrate runs out. The (cyclooctadiene)rhodium(I) complexes, 18 and 19, which are also precatalysts for olefin hydrogenation, show related solvent dependant reactivity. The solvent effects on the reaction of 18 with hydrogen are consistent with the difference in the rate of hydrogenation of octene using complex 15 in toluene and in methanol. Treatment of 18 with hydrogen in toluene results in no reaction after 72 hours, whereas, using the same conditions with methanol as the solvent causes the complex to decompose giv-ing rhodium metal after 1 hour. Analysis of the volatile compounds by GLC revealed that the COD ligand had been hydrogenated completely to cyclooctane; the cyclooctane Chapter 3. Hydrogenation Studies 81 Ik A Figure 3.20: Possible mechanism for the hydrogenation cycle. Chapter 3. Hydrogenation Studies 82 Figure 3.21: *R NMR of bis(7/2-ethene)(l,3-(l-phenyl)butanedionato-0,0 ')rhodium(I) in methanol- d4. Chapter 3. Hydrogenation Studies 83 300-1 T i m e (s) Figure 3.22: Hydrogen uptake plot for 1-octene reduction by hydrogen using complex 15 in toluene. See table 3.12 for the reaction conditions. Chapter 3. Hydrogenation Studies 250 -i 3000 Time (s) Figure 3.23: Ethylene reacted versus time, using complex 15. Chapter 3. Hydrogenation Studies 85 was identified by comparison of the GLC of the reaction mixture with that of pure cy-clooctane. It is unlikely that the COD was hydrogenated homogeneously as attempts to hydrogenate COD with 19 (COD : complex ratio of 50 : 1) in methanol proceeded by initial rhodium precipitate followed by slow hydrogenation. Analysis of the volatile products prior to completion by GLC revealed a mixture of cyclooctadiene, cyclooctene and cyclooctane; again these compounds were identified by comparison of the GLC of the reaction mixture with that of the pure compounds. Attempts to hydrogenate COD with the bis(ethene)rhodium(I) complexes 14 and 15 yielded similar results. Chapter 4 Hydrosilylation Studies 4.1 Introduction The addition of silanes to carbon-carbon multiple bonds and carbon-oxygen double bonds (hydrosilylation) is an important reaction for the production of organosilicon compounds [101] and silyl ethers [102]. Because hydrolysis of the silyl ethers affords the corresponding alcohol the overall process of hydrosilylation of a ketone followed by hydrolysis can be used as a method of reducing ketones to alcohols (equation 4.10). The hydrosilylation of ketones without hydrolysis is also a useful reaction as silyl ethers are important reagents in organic synthesis [103]. (4.10) 0 H H II catalyst / I P h 2 S i H 2 + RCR' • P h 2 S i — 0 — C r - R catalyst: [Rh], [Pt], [Pd] H 3 0 + t 0 I RCR' H 86 Chapter 4. Hydrosilylation Studies 87 (BMPP)2Pt2Cl< (MPPP) 2Pt 2CL, The use of complexes of the platinum metals to catalyze the hydrosilylation of ketones has been studied for a variety of ketones, silanes, ligands and metals [24, 104]. Complexes of rhodium [105], platinum [106], and palladium [107] have been employed as catalysts for this reaction. The first reported example of asymmetric hydrosilylation of ketones was the use of the B M P P and M P P P complexes of platinum for the hydrosilylation of aromatic ketones by methyldichlorosilane giving optical yields of 2.4 - 18.6 % e.e.. [108]. These results preceeded the use of a variety of rhodium(I) complexes, which had been previously employed for the asymmetric hydrogenation of prochiral olefins, for the asymmetric hydrosilylation of prochiral ketones [24]. The proposed mechanism for rhodium catalysed ketone hydrosilylation is outlined in figure 4.24 [23]. The reaction is thought to proceed by the initial formation of a rhodium silyl hydride complex 'A'. In the next step 'A' interacts with a ketone and the reaction follows one of two possible routes. In the first route, the silyl ligand adds to the oxygen of the ketone and the rhodium bonds to the carbon ' B ' . In the second route, the hydride Chapter 4. Hydrosilylation Studies 88 adds to the carbon of the ketone and the rhodium bonds to the oxygen ' C . The next step involves the loss of the silyl ether product from 'B' or ' C , leaving the coordinatively unsaturated rhodium which then reacts with silane regenerating 'A'. The hydrosilylation of ketones by rhodium complexes of chiral phosphorus based ligands have been found to give moderate optical yields for the silyl ether product. The highest optical yield obtained for the 'model' asymmetric hydrosilylation system, the addition of diphenylsilane to acetophenone [24], using these ligands is 74% e.e. using Rh(I)/DIOP [109]. Some results obtained for the model system using chiral phosphorus-based ligands are summarized in table 4.15. Brunner and coworkers have reported asymmetric hydrosilylation catalyst systems giving high optical yields which use nitrogen based ligands 23-28 [115]. The ligands are prepared from readily available optically active starting materials and the catalysts are prepared in situ by the addition of the (cyclooctadiene)rhodium(I) chloride dimer to a solution of ketone, silane and ligand. Generally, an excess of ligand is employed (rhodium : ligand is approximately 1 : 6 ) . Some of the results obtained using this system are summarized in table 4.16. Rhodium complexes of ligands 24-28 were not isolated. The (cyclooctadiene)rhodium(I) complex of 23 has been isolated as the PFg salt and the structure has been verified by x-ray crystallography [57]. The structure shows a square planar cationic rhodium which is bound to the COD olefins and to the imino-pyridine ligand which chelates using both nitrogen donor atoms. Figure 4.24: The proposed reaction mechanism for the catalytic hydrosilylation of ketones using a rhodium catalyst [23]. 00 CD Chapter 4. Hydrosilylation Studies 90 Table 4.15: Hydrosilylation of Acetophenone by Diphenylsilane using Rhodium/Phosphorus-Ligand Systems Ligand Employed Chemical Yield (%) Optical Yield (%e.e.) Reference BMPP 1 84.6 14.6 [39] BMPP 2 98 42 [110] DIOP 3 100 28 [111] DIOP 4 81 30.6 [39] Glucophinite 5 65 55 [112] MPFA 6 85 49.2 [113] BPPFA 7 72 28.6 [114] 1Reaction at 20°C for 40 h, catalyst = 0.1 mol percent, silane/ketone = 1. 2Reaction at 5-25°C for 3-48 h, catalyst = 0.3 mol percent, silane/ketone = 1.1, ketone = PhCOCH 2 CH 3 . 3Reaction at room temperature for 20 h, [catalyst] = 25-40 mM, silane/ketone = 2, ketone/rhodium =50. 4Reaction at 50°C for 40 h, catalyst = 0.1 mol percent, silane/ketone = 1. 5Reaction at room temperature for 12 h, catalyst = 2 mol percent, silane/ketone = 1. 6Reaction at 20°C, catalyst = 0.05 mol percent. 7Reaction at 20°C, catalyst = 0.05 mol percent. Chapter 4. Hydrosilylation Studies 91 MPFA Chapter 4. Hydrosilylation Studies 92 Table 4.16: Hydrosilylation of Acetophenone using Rhodium/Nitrogen-Based Ligand Systems 1 Ligand Employed Time(hours) e.e.(%) Chemical Yield(%) 23 23 51.1 87 24 65 78.2 94 25 57 80.1 93 26 65 9.0 32 27 63 0.4 78 28 24 0.7 19 29 427 0.6 71 30 94 0.1 88 ^odium/ligand = 1/2.9 - 1/20, rhodium/substrate = 1/10 - 1/860, rhodium = 0.1 -0.5 mol percent [115]. Chapter 4. Hydrosilylation Studies 93 H v ^ C H 3 23 ROOC R R' R" 2 4 C H j H H 25 C H 3 C H 3 H 26 H H CH3 27 H 28 C 2 H 5 H. 0 6' yCH3 p-Ph 29 CH3 30 Chapter 4. Hydrosilylation Studies 94 In another paper Brunner and coworkers reported asymmetric ligands prepared from 2- chlorotropolone and methylbenzylamine [116]. Complexes of these ligands with (cy-clooctadiene)rhodium(I) were isolated and characterized. The hydrosilylation of ace-tophenone by diphenylsilane using the in situ generated complexes of 29 and 30 gave low optical yields (table 4.16). 4.2 Results and Discussion The series of bis(ethene)rhodium(I) complexes discussed in chapter 3 as catalysts for the hydrogenation of olefins were also studied with regard to the catalytic hydrosilylation of ketones. Rhodium(I) /3-diketonates have not been reported as catalysts for the hy-drosilylation reaction; most catalysts contain phosphine ligands. The only other known catalysts which utilize hard donor atoms are those studied by Brunner and have been described in the introduction. The bis(ethene)rhodium(I) complexes 13-15 and 17 were tested as catalysts for the hydrosilylation of acetophenone by diphenylsilane. The reactions were performed by the addition of the appropriate complex to a degassed mixture of diphenylsilane and acetophenone. The progress of the reactions were monitored by hydrolysing samples of the reaction mixture and analyzing the products by GLC. The initial results revealed that the system was promising for the hydrosilylation reaction as the chemical yields were reasonable (table 4.17), however, the only optically active complex (17) did not give rise to any measurable optical yield. In order to explore this reaction further some optically active ligands, related to /3-diketonates were prepared. Three new ligands, N-methylbenzylsalicylamide (31), (2'-hydroxy)(l-(R)- a-methylbenzyhmine)ethylbenzene (32), and l-(l-phenyl)(l-(R)- a-methylbenzylai 3- one (33) were synthesized employing the readily available optically active a-(R)-methylbenzylami: Chapter 4. Hydrosilylation Studies 95 Table 4.17: Hydrosilylation of Acetophenone using Bis(ethene)rhodium(I) Complexes 1 , 2 Complex Time (hours) Chemical Yield (%) 13 150 27.4 14 61 83.7 15 150 62.0 17 64 44.7 1 Conversion was monitored by hydrolyzing the silyl ether to the corresponding alcohol which was then analysed by GLC. 2 All reactions were performed at room temperature with a silane to ketone ratio of 105 : 100, catalyst = 0.01 - 1.0 mol percent. as starting material. It was hoped that the presence of the chiral group on the nitrogen donor atom in 31-33 might enhance the ability of the rhodium derivatives to act as asymmetric catalysts. In the rhodium(I) complex of benzoylcamphor, 17, the chiral center is further away from the rhodium metal and therefore may have less effect during the catalysis. Ligand 32 is similar in structure to the nitrogen-based ligand 23, however, some dif-ferences between the ligands are apparent which may influence their effectiveness. Ligand 32 should afford a neutral rhodium(I) complex containing a six-membered heterocyclic ring with oxygen donor atoms; ligand 23 is expected to form a cationic rhodium(I) com-plex with a five membered ring and two nitrogen donor atoms. Another difference is the presence of a methyl group on the imino-carbon of 32 instead of the proton in 23. Chapter 4. Hydrosilylation Studies 96 4.2.1 Preparation and Characterization of the Optically Active Ligands Ligand 31 was prepared by the reaction of salicyl chloride with (R)-a-methylbenzylamine (equation 4.11). The salicyl chloride was prepared by the reaction of salicylic acid and thionyl chloride. (4.11) OH || H OH II V ^ C t ^ c c . + P h _ i _ N H 2 E t z 0 > ^ C H 3 H 31 Ligand 32 was prepared by the reaction between 2'-hydroxyacetophenone and (R)-a-metl^lbenzylamine (equation 4.12). This reaction proceeds with the production of water and its removal facilitates higher conversion; therefore, the conventional Dean-Starke trap technique was employed and the reaction was carried out in refluxing benzene. The reaction was monitored by GLC and it was found that the addition of activated molecular sieves (3A) directly to the reaction mixture facilitated fast conversion to the product. When using the Dean-Starke trap, GLC indicated a conversion of 44% after 24 hours. When molecular sieves were added to the reaction mixture, under the same conditions, a conversion of 99% was observed after 16 hours. Chapter 4. Hydrosilylaiion Studies 97 toluene - H 2 0 (4.12) Ligand 33 was prepared by the reaction of benzoylacetone with methylbenzylamine (equation 4.13). Again, the addition of molecular sieves facilitated the conversion to the product. (4.13) 8 ii ^ C H 3 H 3 C I II , C C H 2 C C H 3 to.uene > - j ^ ^ C C * i ^ N ^ 0 - H 2 0 L V ^ J H 3 3 Chapter 4. Hydrosilylation Studies 98 The chiral ligands were characterized by elemental analysis, mass spectrometry and *H NMR spectroscopy. The elemental analysis (table 4.18), mass spectrometry and : H NMR data (table 4.19) obtained for ligand 3 1 are consistent with its expected structure and formulation. The aromatic protons resonate from 6.8 to 7.5 ppm and have an integral indicative of the nine protons. A doublet at 6.49 ppm (J=8.0 Hz) with an integral indicative of one proton is assigned to proton 'a'; this proton is coupled to one proton, 'b'. A doublet of quartets at 5.31 ppm with an integral indicative of one proton is assigned to proton 'b' which is coupled to protons 'a' and 'c' (J=8.0 Hz). A doublet at 1.63 ppm with an integral indicative of three protons is assigned to the methyl protons 'c'. The hydroxyl proton was not located. The mass spectrum of complex 3 1 contains a peak corresponding to the parent molec-ular ion at 241 m/z. Characteristic fragment peaks include a peak due to the presence of [HOPhCO]+ (120 m/z) and [PhCHCH3]+ (105 m/z). The elemental analysis (table 4.18), mass spectrometry and a H NMR data (table 4.20) obtained for 3 2 are consistent with the expected structure and formulation. A set of multiplets from 6.8-7.5 ppm with ah integral indicative of nine protons is assigned to the aromatic protons. A quartet at 4.95 ppm with an integral indicative of one proton is assigned to proton 'a' which is coupled to the methyl protons 'c' (J=6.7 Hz). A singlet at 2.35 ppm with an integral indicative of three protons is assigned to the 'b' methyl protons. A doublet at 1.64 ppm with an integral indicative of three protons is assigned to the 'c' protons which are coupled to 'a' (J=6.7 Hz). The mass spectrum of 3 2 contains a peak corresponding to the parent molecular ion at 239 m/z and peaks due to characteristic fragment ions. Peaks indicate an ion corresponding to the loss of the hydroxyl group (222 m/z) and ions indicating the presence of [HOC 6 H 4 CNHCH 3 ] + (135 m/z), [HOC 6H 4CNH] + (120 m/z) and [PhCHCH3]+ (105 Chapter 4. Hydrosilylation Studies 99 m/z) fragments. The elemental analysis (table 4.18), mass spectrometry and *H NMR data (table 4.21) obtained for 3 3 are consistent with the expected structure and formulation. A doublet at 11.85 ppm with an integral indicative of one proton is assigned to the amino proton 'a'; this proton is coupled to 'c' (J=7.0 Hz). A set of multiplets from 7.1 to 8.0 ppm with an integral indicating ten protons is assigned to the aromatic protons of the phenyl rings. A singlet at 5.70 ppm with an integral indicative of one proton is assigned to the methyne proton 'b'. A doublet of quartets at 4.77 ppm with an integral indicating one proton is assigned to proton 'c' which is coupled to protons 'a' and 'e' (J=7.0 Hz). A singlet at 1.95 ppm with an integral indicative of three protons is assigned to the methyl protons 'd'. A doublet at 1.61 ppm with an integral indicating three protons is assigned to the methyl protons V which are coupled to proton 'c' (J=7.0 Hz). The mass spectrum of 3 3 contains a peak corresponding to the parent molecular ion at 265 m/z. Fragment peaks are present which correspond to the loss of CH3 (250 m/z), C O C H 3 (222 m/z), and C H 2 C O C H 3 (208 m/z). Peaks are also present which correspond to the [PhCHCH 3NH 2] + (120 m/z) and [PhCHCH3]+ (105 m/z) ions. Chapter 4. Hydrosilylation Studies 100 Table 4.18: Elemental Analyses of the Chiral Ligands Ligand Carbon Hydrogen Nitrogen 31 Expected 74.67 6.26 5.80 Found 74.88 6.44 5.65 32 Expected 80.30 7.16 5.85 Found 80.22 7.06 5.68 33 Expected 81.48 7.22 5.28 Found 81.79 7.34 5.16 Chapter 4. Hydrosilylation Studies Table 4.19: Mass Spectrometry and 1E NMR Data for 31 1 m/z 1 (%) 2 6 (ppm) J(Hz) Assignment Protons 3 241 9.4 6.8-7.5 aromatic 9 137 56.5 6.49(doublet) 8.0 a 1 120 40.8 5.31(doublet of quartets) 8.0 b 1 105* 100.0 1.63(doublet) 8.0 c 3 1 The NMR data was obtained at room temperature using C D C I 3 as solvent. 'Intensity (1) is calculated relative to the base peak (*). 3 Tlie number of protons is calculated by relative integral intensity. 31 Chapter 4. Hydrosilylation Studies Table 4.20: Mass Spectrometry and a H N M R Data for 32 1 m / z I (%) 2 £(ppm) J(Hz) Assignment Protons 3 239 40.4 6.8-7.5 aromatic 9 222 8.7 4.95(quartet) 6.7 a 1 148 11.3 2.35 b 3 135 40.4 1.64(doublet) 6.7 c 3 120 .21.9 105* 100.0 •"The N M R data was obtained at room temperature using C D C 1 3 as solvent. in tens i ty (I) is calculated relative to the base peak (*). 3 T h e number of protons is calculated by relative integral intensity. 32 Chapter 4. Hydrosilylation Studies Table 4.21: Mass Spectrometry and 'H NMR Data for 33 1 m/z I (%) 2 £(ppm) J(Hz) Assignment Protons 3 265 23.5 11.85(doublet) 7.0 a 1 250 4.1 7.1-8.0 aromatic 10 222 11.5 5.70 b 1 208 4.2 4.77(doublet of quartets) 7.0 c 1 167 4.2 1.95 d 3 160 38.9 1.61(doublet) 7.0 e 3 145 5.8 120 11.0 105* 100.0 •'The NMR data was obtained at room temperature using C D C I 3 as solvent. 'Intensity (I) is calculated relative to the base peak (*). 3The number of protons is calculated by relative integral intensity a c / H a I N 0 c H 3 C I II d Or V & C C H 3 3 3 Chapter 4. Hydrosilylation Studies 104 4.2.2 Asymmetric Hydrosilylation Studies The chiral ligands were tested for the in situ catalytic hydrosilylation of acetophenone by diphenylsilane. Two methods were devised for this reaction, the first involved adding the bis(ethene)rhodium(I) chloride dimer to a solution of diphenylsilane, acetophenone and the free ligand. The second method devised was to add the bis(ethene)rhodium(I) chlo-ride dimer to a solution of diphenylsilane acetophenone and the sodium salt of the ligand. The first method is similar to that used by Brunner for the asymmetric hydrosilylation of ketones by diphenylsilane using ligands 23-30 and the (cyclooctadiene)rhodium(I) chlo-ride dimer as the rhodium source [115, 116]. The second method was devised because of the success in preparing the rhodium(I) complexes discussed in .Chapter 3 using the sodium salts of the ligands. When the second method was attempted with 32 the planned protocol was abandoned after the sodium salt of the ligand was added due a vigorous reaction resulting from its addition. Further investigation revealed that the product of the reaction was the silyl ether expected from the rhodium catalysed reaction. This novel and unexpected reaction was studied to determine its utility for a variety of ketones and its selectivity was tested using 2-cyclohexen-l-one; the results from these studies are presented in the appendix. The results obtained for the hydrosilylation of acetophenone using the first method are shown in table 4.22. Although the use of these catalysts resulted in reasonable chemical yields, their optical yields were disappointing. It was hoped that the reaction, with the catalyst generated from ligand 32 and the rhodium dimer, would give high optical yields because of its structural similarity to 23 but the differences must have been crucial. Ligand 33 affords the highest optical yield; this result surpasses the optical yields reported by Brunner et al for chiral ligands with both oxygen and nitrogen donor atoms (table 4.16, 29 and 30). This optical yield is low, however, in comparison to the Chapter 4. Hydrosilylation Studies 105 Table 4.22: Hydrosilylation of Acetophenone by Diphenylsilane using Chiral Oxygen/Nitrogen Donor Ligands1,2,3 Ligand Time(hours) Chemical Yield(%) Optical Yield (%e.e.) 31 142 83.4 2.39 32 67 88.0 0.04 33 52 96.2 8.65 1 Conversion was monitored by GLC. Conversions were calculated by the relative amounts of ketone and alcohol present after hydrolysis of the silyl ether. 2Enantiomeric excess was calculated from optical rotation measurements. 3AU reactions were performed at room temperature with a silane to acetophenone ratio of 105 : 100, rhodium to ligand ratio of 1 : 5, rhodium = 0.03 - 0.8 mol percent. rhodium catalyst sytems employing phosphorus- or nitrogen-based ligands. Among the possible reasons for the low optical yields is that the chiral centers in the ligands are too far from the ketone substrate during the silane insertion step. Bis(ethene)rhodium(I) complexes of the ligands were not obtained in their pure form, nor were crystals suitable for x-ray study, so that one can only speculate on the distance between the chiral center and the metal. The crystal structure of ligand 31 was determined by Steve Rettig of the Univer-sity of British Columbia and is shown in figure 4.25. The amino hydrogen is hydrogen bonded to the hydroxyl oxygen forming a six membered ring. The rhodium atom in the corresponding bis(ethene)rhodium(I) complex may be found in the position of the amino proton or may be bound by the phenoxy oxygen and either the ketone oxygen or the nitrogen atom; the two latter modes of bonding -would allow for the formation of six-7r-electron heterocycles similar to the rhodium diketonate complexes. Figure 4.25: Crystal structure of N-methylbenzylsalicylamide, 31. Chapter 4. Hydrosilylation Studies 107 H \ /NMe2 FA 4.2.3 Asymmetric Ligands Based on a-N,N-Dimethylaminoethylferrocene The elements of chirality present in ligands prepared from N,N-dimethylaminoethyl-ferrocene, FA, are known to be effective for the induction of chirality into the products of catalytic hydrogenations [117]. Some catalysts employing these ligands have been shown earlier to be useful for asymmetric hydrosilylation reactions, affording moderate optical yields (table 4.15). The use of FA-based ligands in asymmetric catalysts was prompted by the work of Ugi et al who reported that FA can be easily resolved into its two antipodes and that lithiation of FA occurs stereospecifically [118]. Substitutions on FA performed via the lithiation reaction give rise to a product which has a chiral plane, and the stere-ospecific nature of the lithiation reaction dictates the configuration of the chiral plane (figure 4.26). Ligands prepared in this way from (-R)-FA have predominantly (R,S)-configuration and those prepared from (5)-FA have predominantly (5, .Reconfiguration. Several ferrocenylphosphine ligands have been prepared by this route using the reaction of lithio-FA with alkyl- and aryl-substituted chlorophosphines [117, 119]. An attempt was made to make use of the FA chirality and lithiation directing ability by preparing a diketone ligand of the type shown in figure 4.27. The synthetic route cho-sen to prepare this ligand is shown in figure 4.28 and proceeds via the initial preparation of 2'-(a-N,N-dimethylaminoethylferToCene)carboxyhc acid, FACOOH. Chapter 4. Hydrosilylation Studies 108 Figure 4.26: Production of planar chirality in FA derivatives. Figure 4.27: Target ligand based on FA. Chapter 4. Hydrosilylation Studies 109 NMe 2 Figure 4.28: Synthetic route for the FA based diketone ligand. The FACOOH was prepared by the addition of solid carbon dioxide to a solution of lithio-FA (prepared by the addition of butyllithium to a solution of FA) and was isolated as the lithium salt of the carboxylic acid (88% yield). The salt was converted to the free acid by hydrolysis. This reaction is analogous to the preparation of mono- and 1,1'-ferrocenedicarboxylic acids [120]. Esters of the FACOOH were prepared by the reaction of the acid with dialkylsulfates. The methyl and ethyl esters were prepared in this manner with yields of 49% (methyl) and 60% (ethyl). A condensation reaction between the esters and deprotonated ketones was then attempted. This reaction is similar to the one used to prepare the 1,3-dioxobutylferrocene ligand as well as benzoylcamphor (chapter 3); unfortunately, it did not work. Attempts were made for the reaction of both the methyl and ethyl esters of the FACOOH with the lithium salts of acetophenone, acetone, and acetylferrocene with no indication of a reaction (the reactions were monitored by thin layer chromatography). Further aspects of these reactions are discussed in the appendix. Chapter 4. Hydrosilylation Studies 110 The FACOOH and its methyl and ethyl esters were characterized by the use of ele-mental analyses (table 4.23), mass spectrometry and *H NMR spectroscopy. The elemental analysis data for FACOOH gave values which were outside the limits of experimental error for the pure complex, however, the obtained elemental analysis data matches the formulation of the hemihydrate of FACOOH. Mass spectrometry and 1U NMR data (table 4.24) are consistent with the expected structure and formulation. Complexes prepared from the FACOOH were isolated in pure form and their characteri-zation further supports the formulation of FACOOH. The *H NMR spectrum at room temperature is consistent with the expected structure, however, broadening of the resonances is present due to the fluxional effects. A variable temperature *H NMR study of FACOOH is discussed in the appendix. A broad resonance at 17.7 ppm with an integral indicative of one proton is assigned to the hydroxyl proton. The high shift may be due to hydrogen bonding effects, discussed in more detail in the appendix. A multiplet at 5.01 ppm with an integral indicating one proton is assigned to proton 'b' of the substituted ferrocenyl ring due to its proximity to the deshielding carbonyl group. A quartet at 4.53 ppm with an integral indicating one proton is assigned to proton 'c' which is coupled to the methyl 'h' protons (doublet at 1.40 ppm) with a coupling constant of 6.5 Hz. Overlapping multiplets at 4.28 ppm with an integral indicative of two protons are assigned to the 'd' protons of the substituted ferrocenyl ring. A singlet at 4.10 ppm with an integral indicating five protons is assigned to the protons of the unsubstituted ferrocenyl ring. Two broad singlets at 2.54 and 2.03 ppm with integrals indicative of three protons are assigned to the N-methyl protons T and 'g'; these resonances coalesce at 38.5°C. The mass spectrum of FACOOH contains a peak corresponding to the parent molec-ular ion at 301 m/z. Fragment peaks are present due to the loss of a methyl group (286 m/z), loss of the COOH group (256 m/z), and loss of the dimethylaminoethyl group (230 Chapter 4. Hydrosilylation Studies 111 m/z). A peak at 186 m/z is present due to the ferrocene ion. The *H NMR spectrum of the methyl ester of FACOOH, FACOOMe, is similar to that of FACOOH (table 4.25). A singlet due to the methyl group is present at 4.35 ppm. The protons of the N,N-dimethylamino groups appear as a singlet at 2.06 ppm. The mass spectrum of FACOOMe (table 4.25) contains a peak due to the presence of the molecular ion at 315 m/z as well as fragments due to the loss of C H 3 (300 m/z), HNMe2 (270 m/z) and subsequent loss of OCH 3 (239 m/z). The ethyl ester of FACOOH, FACOOEt, has a similar *H NMR spectrum to that of FACOOMe (table 4.26). The differences arise from the two methylene protons which appear as a quartet in the ferrocenyl region (4.2-4.36 ppm); this resonance is obscurred by overlap with the protons of the substituted ferrocenyl ring. The methyl group gives rise to a triplet at 1.34 ppm (J=6.9 Hz). The mass spectrum of FACOOEt contains a peak due to the presence of the parent molecular ion (329 m/z) as well as peaks due to the fragment ions from the loss of C H 3 (314 m/z), NMe2 (285 m/z) and COOCH 2 CH 3 (256 m/z). Chapter 4. HydTosilylation Studies Table 4.23: Elemental Analyses Data for FACOOH and its Esters Complex Carbon Hydrogen Oxygen Nitrogen FACOOH Expected 59.82 6.36 4.65 Expected 1 58.18 6.51 4.52 Found 58.31 6.89 4.58 FACOOMe Expected 60.97 6.72 4.44 Found 61.13 6.81 4.61 FACOOEt Expected 62.02 7.04 4.25 Found 61.73 7.00 4.06 1 Calculated values for FACOOH • 1/2 H 2 0 Chapter 4. Hydrosilylation Studies 113 Table 4.24: Mass Spectrometry and J H NMR Data for FACOOH 1 m/z I (%) 2 £(ppm) Assignment Protons 3 301 36.05 17.7 a 1 286 21.10 5.01 b 1 256* 100.0 4.53(quartet,6.5 Hz) c 1 242 37.21 4.28 d 2 230 19.63 4.10 e 5 212 60.93 2.54 f 3 186 11.86 2.03 g 3 1.40(doublet,6.5 Hz) h 3 aThe NMR data was obtained at room temperature using CDC13 as solvent. 'Intensity (I) is calculated relative to the base peak (*). 3The number of protons is calculated by the relative integral intensity. Chapter 4. Hydrosilylation Studies 114 Table 4.25: Mass Spectrometry and : H NMR Data for FACOOMe 1 m/z I (%) 2 6(ppm) Assignment Protons 3 315 0.65 4.75 a 1 300 1.88 4.43(quartet,7.0 Hz) b 1 270 1.62 4.35 c 2 239 1.00 4.18 d 5 59* 100,0 3.79 e 3 2.06 f 6 1.49(doublet,7.0 Hz) g 3 1The NMR data was obtained at room temperature using CDC13 as solvent. intensity (I) is calculated relative to the base peak (*). 3The number of protons is calculated by the relative integral intensity. Chapter 4. Hydrosilylation Studies 115 Table 4.26: Mass Spectrometry and J H N M R Data for F A C O O E t 1 m / z I (%) 2 6 {ppm) Assignment Protons 3 329 33.82 4.78(multiplet) a 1 314* 100.00 4.46(quartet,6.9 Hz) b 1 285 15.87 4.08, 4.2 - 4.36 c, d 4 263 20.17 4.13 e 5 256 17.41 2.07 f 6 239 67.97 1.49(doublet,6.9 Hz) g 3 211 11.01 1.34(triplet.6.9 Hz) h 3 ! T h e N M R data was obtained at room temperature using CDCI3 as solvent. i n t e n s i t y (I) is calculated relative to the base peak (*). 3 T h e number of protons is calculated by relative integral intensity. NMe 2 / d ft CH2CH3 Chapter 5 N M R Spectroscopic Studies 5.1 Introduction The study of fluxionahty in organometallic molecules is useful as it can afford information about the strength and type of interaction between ligand and metal, steric interactions between ligands, bond hybridizations, preferred geometries, and can lead to predictions of reactivity. NMR studies have revealed a variety of fluxional processes and techniques are available to correlate exchanging nuclei and, more importantly, determine rates of exchange [47, 53, 121, 122, 123, 124]. Determination of the exchange rates at a range of temperatures allows the calculation of the activation parameters for the fluxional processes. Data for the simplest calculation of AG* can be acquired from the NMR spectrum at the temperature at which the peaks due to the exchanging nuclei coalesce. By substitut-ing the temperature at the coalescence point and the peak separation at slow exchange into equation 5.14 it is possible to obtain values for AGK AG* = -RT\n^± (5.14) AG* = Gibb's free energy of activation. R = Gas constant. T = Temperature of coalescence. 116 Chapter 5. NMR Spectroscopic Studies 117 Ai/ = The peak separation in the absence of exchange in hertz. h = Planck's constant. k' = Boltzmann's constant. A useful, and more accurate technique for determining AG*, first described by Forsen and Hoffman [125], is spin saturation transfer. In their method a resonance absorption due to one set of the exchanging nuclei is saturated and the decrease in intensity of the resonances due to the nuclei with which it is exchanging are monitored. Once the system has reached a steady state, during saturation of one of the exchanging nuclei, the magnetism of the exchanging nuclei which are not being irradiated is dependent on the initial magnetism, the longitudinal relaxation rate (Ti), and the rate of exchange. Measurement of the initial and steady state magnetizations can be made by integration of the resonance or by peak height. The T\ for the nuclei can be measured by using conventional techniques and will be discussed later. The exchange rate k can then be calculated from equation 5.15. For an intramolecular process, the exchange rate equals the rate constant. k = The exchange constant governing the exchange between the observed nuclei and the irradiated nuclei. Ti = Longitudinal relaxation time (seconds). Mi = Initial magnetization of the observed nuclei. Mz = Steady state magnetization of the observed nuclei during saturation of the irradi-ated nuclei. Chapter 5. NMR Spectroscopic Studies 118 The activation parameters can be calculated from the plot of ln(fc/T) versus 1/T1. The enthalpy of activation Aif* is obtained from the slope by equation 5.16. The entropy of activation AS* is obtained from the Y-intercept by equation 5.17. The free energy of activation can then be calculated at temperature T by equation 5.18. AHX = (1.987 x 1(T3) • (slope)(Kml/mol) (5.16) AS* = (1.987 x 1CT3) • (Yintercept - 23.76){Kcal/mol • K) (5.17) AG* = A # * - T A S * (5.18) A disadvantage of this spin saturation transfer technique is that it is complicated when the exchanging nuclei are coupled. This complication arises because of the presence of the nuclear Overhauser effect (NOE). In systems in which NOE enhancement is expected, it is necessary either to know the exact value of the enhancement factor or to turn to another method such as computer fitting. The use of dynamic NMR fitting programs give reasonably accurate values of k which can then be used to calculate the activation parameters. The fitting programs are em-ployed by first inputting values describing the exchanging system such as the coupling constants, transverse relaxation time (T2), and the chemical shift difference for the sets of exchanging nuclei. Values of k are then entered and the program generates a theoret-ically calculated spectrum from the entered parameters. When the simulated spectrum matches the observed spectrum the values of k are used to calculate the activation pa-rameters from the plot of \n(k/T) versus 1/T as before. An advantage of the NMR fitting technique is the ability to determine exchange rates which are much higher than those r^om the relation k - (k'T/h)e^ Chapter 5. NMR Spectroscopic Studies 119 which are accessible by the use of spin saturation transfer. The spin saturation transfer method is unable to determine high exchange rates because M 2 becomes zero and equa-tion 5.15 requires non-zero M 2 values. The fitting techniques are capable of determining exchange rates as long as line broadening (due to the exchange) is present. One of the disadvantages of the dynamic NMR fitting programs is that the exchang-ing system must be well understood. The technique in itself affords little information regarding the exchange and it is possible to generate a spectrum resembling the observed spectrum by using improper assumptions regarding the exchange process. Other methods for calculating k, which are based on the measurement of the rate of change of magnetization in the exchanging system after saturating one of the exchanging nuclei, have been described by Noggle and Shirmer [53] and will be presented in further detail later. Dynamic NMR studies of coordinated olefins have revealed that these ligands are often involved in fluxional processes [126, 127]. The 1 H NMR spectra of bis(ethene)rhodium(I) complexes exhibit fluxionality associated with the ethylene ligands. Cramer [55] described the first examples of this class: bis(7;2-ethene)(2,4-pentanedionato-0,0 ')rhodium(I) 1 and bis(7/2-ethene)(7/s-cyclopentadiene)rhodium(I) 2 (figure 5.29, R=C 2H 4). He attributed the fluxionality as due to the rotation of the ethylene ligands about the ethylene-rhodium bond axis. Cramer also observed exchange between the ethylene ligands of 2 and gaseous ethylene. This exchange was present at temperatures as low as — 78°C. A similar ex-change between [(C2H4)2PtCl2]2 and gaseous deuterated ethylene at — 15°C has also been reported [128]. Ethylene rotational barriers in the complexes of structure 1 vary with R as follows (R, Ea(Kcal/mol)) : C 2 H 4 , 15.0; S0 2 , 12.2; C 2 F 4 , 13.6 [54]. The lower rotational barriers of ethylene in the sulfur dioxide and tetrafluoroethylene complexes were attributed to weakening of the rhodium-ethylene bonds due to the electron withdrawing nature of the Chapter 5. NMR Spectroscopic Studies 120 H > S H / M \ H A H Figure 5.29: Structure of bis(7;2-ethene)M complexes. substituent ligands. This causes a decrease in the electron density on the rhodium metal, lowering the amount of backbonding with the olefin. Many related rhodium(I) complexes have been studied by NMR techniques since Cramer's initial report. Some of the results related to the systems studied in this thesis are summarized next. A series of bis(772-ethene)(4-imino-2-pentanato-0,N)rhodium(I) complexes (34) (R = H, C H 3 , Ph, C 6 H 4 C H 3 A H 4 C I , C 6H 4Br, C 6 H 4 O C H 3 , C 6 H 4 N0 2 ) were studied by Kriz the 1 H NMR spectrum and by study of the 1 H NMR of carbon monoxide substution reaction products it was concluded that the ethylene trans to the nitrogen was more labile, presumably due to the trans effect. A study with similar implications to those of Kritz and Bouchal was performed by Miya and Santo using platinum complexes [130]. They correlated the trans influence of and Bouchal [129]. Non-equivalence of the fluxional ethylene ligands was apparent in Chapter 5. NMR Spectroscopic Studies 121 Ik A p Rh 34 Br Cl 35 a varietj' of pyridine derivatives to the AG* for ethylene rotation in the position trans to the heterocyclic ligand in the bromochloro(7/2-ethene)(4-R-pyridine-N)platinum(II) com-plexes (35) (in order of increasing AG 1 : R = Cl, CN, CH 3 , H). The measured AG 1 for ethylene rotation decreased when the substituted pyridine trans to the ethylene had a lower pK„. Again, the lower barriers to rotation were attributed to lower electron density on the metal center. Values for AG* ranged from 14.3 Kcal/mol to 16.0 Kcal/mol. Maisonant and Poilblanc studied the bis(r/2-ethene)(2,6-lutidine-N)rhodium(I) chlo-ride complex 36 which exhibits a 1 H NMR spectrum indicative of two chemically and magnetically inequivalent ethylene ligands [131]. Study of this and other complexes lead them to suggest the possibility of a 'gear mechanism' which couples the motion of the ethylene ligands allowing them to rotate only at similar rates. A variety of bis(772-olefin)(2,4-pentanedionato-0,0 ') rhodium(I) complexes were pre-pared and their fluxionaUty studied by Herberhold et al [132]. The calculated values of Chapter 5. NMR Spectroscopic Studies 122 AG* for olefin rotations ranged from.11.7 Kcal/mol for the ethylene ligands in 37 to 18.5 Kcal/mol for the tetramethoxyethylene ligand in 38. The authors also invoked a 'Berry pseudorotation' to account for the coalescence of the resonances of the pentanedione methyl groups at higher temperatures in 38 and 39. Values of AG* for the motion were calculated to be 13.9 Kcal/mol (38) and 15.2 Kcal/mol (39). A schematic representation of the fluxional process for the four-coordinate complexes is shown in figure 5.30. Bis(ethene)rhodium(I) complexes of Schiff bases have been studied by Mague and Nutt [69]. They reported the *H NMR spectrum of bis(T/2-ethene)(N-methylsahcylaldimine-0,N)rhodium(I), which exhibited fluxionality due to the ethylene moieties. At low tem-perature it is evident that the ethylenes experience different environments in the molecule as four different sets of resonances (twa per ethylene) are present in the 1B. NMR spec-trum. Activation parameters for the ethylene rotations were not reported. Ethylene rotation barriers are higher for iridium complexes than for rhodium. This Chapter 5. NMR Spectroscopic Studies 123 B Rh cf N B B Rh Rh 0/ »0 Figure 5.30: Schematic representation of the proposed Berry pseudorotation for 4-coor-dinate rhodium complexes. observation was reported by Maitlis and coworkers as a result of the study of (T; 6-pentamethylcyclopentadiene)rhodium(I) and iridium(I) ethylene complexes [133, 134, 135]. The ethylene rotation barriers for platinum complexes are similar to those of rhodium as calculated values of A G 1 for ethylene rotation on platinum range from 10-20 Kcal/mol [136, 137, 138, 139, 140, 141, 142, 143]. Five coordinate complexes of d8 metals can undergo concerted ligand rotation and Berry pseudorotation rearrangements [143]; examples of this type of fluxionahty with ethylene or another olefin as one of the ligands have been studied by X H NMR spec-troscopy. Bis(isocyanide)( 772-tetracyanomethylene)(2,4-pentanedionato-0,0 ')rhodium(I) complexes of general structure 40 were studied by Kaneshima et al for a variety of R groups Chapter 5. NMR Spectroscopic Studies 124 CH 3 R C R C N a=CN N R R 40 40a (R= t-Bu, p-CH 3 OC 6 H 4 , p -CH 3 C 6 H 4 , p-ClC 6 H 4 , 2,4,6-(CH3)3C6H2) [144]. The acti-vation energy of rotation of the TCNE ligand coupled with a Berry pseudorotation was calculated to be 15.0 Kcal/mol. The related bis(isocyanide)(olefin)(dimethylthio-carbamato)rhodium(I) complexes were studied by *H NMR spectroscopy and calculated AG* values for the coupled rotation and Berry pseudorotation process ranged from 10.9 Kcal/mol (R' = p-CH 3 C 6 H 4 , olefin = maleic anhydride) to 18.4 Kcal/mol (R' = 2,4,6-(CH3)3C6H2, olefin = tetracyanoethylene) [145]. The mechanism of the Berry pseudoro-tation in five coordinate rhodium complexes is shown in figure 5.31. The concerted olefin rotation and Berry pseudorotation has also been observed in iron complexes. Takats and coworkers reported activation energies varying from 11.1 Kcal/mol (41) to 14.3 Kcal/mol for the fluxional process in 42a [146]. Osborn et al reported a similar rearrangement for a variety of iron carbonyls and estimated AG* to Chapter 5. NMR Spectroscopic Studies 125 Figure 5.31: Berry pseudorotation for 5-coordinate rhodium complexes. be 13 Kcal/mol for 42a,b and greater than 15 Kcal/mol for 43 [147]. Olefin rotational barriers have been studied theoretically for a number of molecular conformations [143, 148, 149]. An in-depth look at the bonding between ethylene and square planar platinum(II) in KPtCl3(C2H 4) indicated that the barriers to rotation are best estimated when the ethylene hydrogens are bent away from the metal to limit steric interactions with the other ligands [143]. Albright and coworkers also concluded that the resting position of ethylene (perpendicular to the plane of the molecule) is determined by steric factors rather than electronic. This conclusion is supported by the observation that most square planar carbene complexes adopt this same geometry although their p-orbital (similar in function to the ethylene 7r*-orbital in accepting back donation) would dictate a 'wrong' conformation if it were to interact with the lower lying 62 orbital; instead, the p-orbital interacts with the higher energy 6a orbital to give the conformation similar to that of the ethylene complex [143] Chapter 5. NMR Spectroscopic Studies 126 0 c olefin 41 H 2 C=C(COOCH 2 CH 3 ) 2 C 0 42 a) <ran5-CH(COOCH2CH3)=CH(COOCH2CH3) 42 b) trans-CH(COOCH3)=CH(COOCH3) 43 C12C=CF2 A number of five coordinate complexes containing chelating diolefins, acetylaceto-nate, and dithiocarbonates, which are seen to undergo the rotation/Berry pseudorota-tion mode of fluxionality have been reported [144, 145, 150]. Studies employing X H NMR spectroscopy have allowed the calculation of activation energies for the motion yielding values for the free energy of activation ranging from 10 Kcal/mol to 15 Kcal/mol. . The theoretically calculated activation energy for ethylene rotation in five-coordinate trigonal bipyramidal d8 complexes is 32 Kcal/mol; when the concerted pseudorotation is added, the estimated activation energy for the entire process drops to 10 Kcal/mol [143]. This value is in agreement with the previously discussed results. The 'Berry pseudorotation' rearrangement has also been proposed to account for the fluxional *H NMR spectral behavior for some four coordinate rhodium complexes with acetylacetonate (discussed earlier) and norbornadiene ligands [132, 151, 152, 153]. The-oretical calculations for this fluxional process in four-coordinate complexes of d8 metals Chapter 5. NMR Spectroscopic Studies 127 have not yet been reported. 5.2 Results and Discussion The bis(ethene)rhodium(I) complexes studied for catalytic abilities in Chapters 3 and 4 reveal fluxionality in their X H NMR spectra due to the ethylene ligands. As there seems to be a relationship between the lability of the olefin ligands and the catalytic activity of the complexes, a NMR study was undertaken to gain insight into the nature of the non-rigidity. Complex 15 was selected as the most suitable for a detailed NMR study as at low temperatures the ethylenes give rise to well separated resonances in the 1 H NMR spec-trum. Cramer noted that at low temperature the X H NMR spectrum of the symmetrical complex bis(r72-ethene)(2,4-pentanedionato-0,0 ')rhodium(I) gives rise to two separate resonances due to the ethylene ligands. The high field resonance was assigned to the 'in-ner' ethylene protons (i) and the low field resonance was assigned to the 'outer' ethylene protons (o) (figure 5.32) [55]. The J H NMR spectrum of 15 at -40°C (figure 5.33) reveals two sets of inner and outer protons, one set per ethylene, attributable to the asymmetric nature of the 2'-hydroxyacetophenone ligand. The ethylene nearest to the deshielding ketone function-ality is expected to be shifted further downfield, separating the ethylene resonances. Another possible rationale for the separation is the observation that the more strongly bound an ethylene ligand is to the metal center, the higher the resonance will be shifted upfield [54]. In complex 15 it is expected that the ketone oxygen exerts a lesser trans effect than the phenoxide oxygen [154]. It would then be expected that the ethylene trans to the ketone would be more strongly bound and would appear further upfield in the NMR spectrum. Both of these rationales suggest that the ethylene cis to the ketone Chapter 5. NMR Spectroscopic Studies 128 Figure 5.32: Diagram of a bis(ethene) rhodium(I) complexes showing 'inner' (i) and 'outer' (o) protons. (ethylene I) would be the downfield ethylene and the ethylene trans to the ketone, the upfield ethylene (ethylene II). At room temperature all of the ethylene protons give rise to a single broad resonance. The coupling patterns of each ethylene can be described as AA'XX'. Assignment of the protons to particular ethylene ligands was made on the basis of decoupling ex-periments; since the different sets of ethylene protons belonging to a single ethylene are coupled, saturating one of the sets should remove the coupling from the other set on the same ethylene. The assignment of the AA'XX' sets on each ethylene was made as follows: irradiating the A resonance at 4.05 ppm (AA' of ethylene I) decouples the C resonance at 2.35 ppm (XX' of ethylene I); irradiation of the B resonance at 3.63 ppm (AA' of ethylene II) decouples the D resonance at 2.18 ppm (XX' of ethylene II) (figure 5.34). Irradiation of one of the resonances, such as A, also causes a decrease in intensity of the resonance due to the protons on the same ethylene, C , aside from the decoupling Chapter 5. NMR Spectroscopic Studies Chapter 5. NMR Spectroscopic Studies 130 -48.5°C 4.6 4.0 3.4 23 22. ppm Figure 5.34: Decoupling the A protons (top) and decoupling the B protons (bottom). effect. This is due to spin saturation transfer. This effect is apparent at temperatures above -60°C. Irradiation of A or C also causes a decrease in the intensity of B and D, which are associated with the other ethylenes (a decrease in A and C is also noted upon the irradiation of B or D). This saturation transfer effect is less pronounced than that caused by exchange between protons of the same ethylene and was only apparent at temperatures above -45°C. Two mechanisms are necessary to account for these observations. Rotation of the ethylene ligands about the rhodium-ethylene bonds exchanges nuclei sets A and C, and B and D, accounting for the first type of saturation transfer. Dissociation of ethylene seems to be necessary to account for the second mode of exchange. 5.2.1 Ethylene Rotation As mentioned earlier the spin saturation transfer method first described by Forsen and Hoffmann is often used for studying the exchange in fluxional molecules. In the absence Chapter 5. NMR Spectroscopic Studies 131 of NOE effects, this experiment can be applied for the measurement and calculation of the exchange rate k for olefin rotational processes. However in the case of 15 the coupling interaction in the ethylene ligands implies that NOE will be present during the decoupling of one of the sets of resonances. If NOE is present, the observed resonance will have a higher intensity than expected for a given exchange rate during spin saturation transfer. The presence of NOE will then cause calculated k values to be artificially low. The amount of NOE enhancement present is difficult to quantify, as the longitudinal relaxation rates of the ethylene protons are temperature dependent [122]. The Ty values for the protons on the ethylene ligands of 15 were measured at a variety of temperatures by using the inversion recovery method; the pulse sequence is shown in figure 5.35. Plots illustrating the variance of Ty with temperature can be seen in figure 5.36 (A and C nuclei) and figure 5.37 (B and D nuclei). These plots match the shape of the theoretically predicted TVtemperature dependance plot. An example of the observed spectra collected at increasing values of D2 is shown in figure 5.38. The Ty values were calculated from the collected spectra using an exponential fit program; Ty equals the slope of the linear regression line from the plot of ]XL(I/I00') versus time. Alternative approaches to Forsen and Hoffmann's method for the calculation of k are described by Noggle and Shirmer [53]. These methods offer a means of measuring spin saturation transfer before NOE has made an effect on the magnetization or relaxation of the spin system. The rate constant for the exchange is calculated by two experiments: the first (experiment 1) yields the value of the total direct relaxation rate R and the second (experiment 2) yields a value which is the difference between R and the rate for the exchange. In experiment 1 the total direct relaxation parameter R is measured by the rate of decrease in magnetization of a resonance after instantaneous saturation of the resonance with which it is exchanging. This is achieved experimentally by using the pulse sequence Chapter 5. NMR Spectroscopic Studies 132 180* 90* F I D Figure 5.35: The pulse sequence employed for the measurement of T^. shown in figure 5.39. An example of experiment 1 is shown in figure 5.40, where the resonance at 4.1 ppm has been saturated and the spectrum collected after set time intervals after saturation. The system is described by equation 5.19 and the value of Ri is calculated from the slope of the plot of ln(/; — I^) versus time (Ri — —slope). The time allowed before data collection (D2) varies from 0 to a value which is less than the time required for NOE to build up, this period is illustrated in the plot of ln(7 — loo) versus time (figure 5.41). This plot represents the effect of the instantaneous saturation on the set of nuclei with which the irradiated nuclei are exchanging at specific time intervals after initial saturation. The experiment was performed at —38.5°C. The initial slope 'a' is equal to —R and the second slope 'b' reflects a combination of — R and NOE. <IzA >=IoA + (^-)IoB(l-e-RAt) (5.19) < IzA > = The magnetization of nuclei A, as measured in the z direction, at time t. IOA = The equilibrium magnetization of nuclei A. Chapter 5. NMR Spectroscopic Studies Figure 5.36: The temperature dependence of Ti for A and C ethylene protons Chapter 5. NMR Spectroscopic Studies Figure 5.37: The temperature dependence of T a for B and D ethylene protons. Figure 5.38: 7\ experiment at -38.5°C, ethylene protons of 15 at i increasing D2. CO Cn Chapter 5. NMR Spectroscopic Studies 136 180' 90* DECOUPLER ON Figure 5.39: The pulse sequence employed for the determination of R. JOB = The equilibrium magnetization of nuclei B . a A ~ The cross relaxational parameter measured from nuclei A during the irradiation of nuclei B . RA — The total relaxation parameter of nuclei A. t = time(seconds). Values of the cross relaxational time ((TAB) {^AB = — ^exchange) in the absence of NOE are obtained from the equilibrium magnetization of the observed spin which equals IOA +IOB{O'AB/RA)- However, in 15 the equilibrium magnetization reflects both CAB !RA and NOE so that k cannot be acquired from this experiment alone. Having determined values of R, from experiment 1, the exchange rates are then obtained by performing experiment 2. This experiment affords values for the sum of R+er for the ethylene protons. Here one of the resonances is irradiated until the steady state is reached. The decoupler is then turned off and the resonances are monitored as they approach their equilibrium magnetizations. This is achieved experimentally by using the pulse sequence shown in figure 5.42. The system is now described by equation 5.20 [54]. At higher exchange rates the slope of hi^I^A ~ IOA) versus time equals RA + <?AB Chapter 5. NMR Spectroscopic Studies 138 Chapter 5. NMR Spectroscopic Studies 139 90* FID DECOUPLER ON Figure 5.42: The pulse sequence employed for the determination of <r. (from equation 5.20) as RA — (rAB is very small compared to RA + O~AB-< IZA >= IOA + Cie-Xlt + C2e~X2t (5.20) *OA\ A j _ A ] ; -r *OA R B V A 2 - A J > Cn — —In A ( ZAILIZZJiA- \ — /_ . ££A ( —hi— \ ^2 *OA\ \ 7 - \ L ) JOA Rj3 K x2-x1 J A: = \{{RA + RB) + [{RA ~ RB)2 + ^AB^BA}'} A2 = ${(RA + RB) - [{RA - RB)2 + 4<xABaBA\^} where RA * RB and o~AB — &BA , Ai = R + tr and A2 = R — o~ The cross relaxational constant (—k) is obtained by the difference between the slopes of the plots in experiment 1 and experiment 2 (ITAB = (RA + crAB (experiment 2)) — (/^ (experiment 1)) = —A:). The activation parameters AH*, AS*, and AG* values are then calculated from the plot of ln(fc/T) versus 1/T. Chapter 5. NMR Spectroscopic Studies 140 Table 5.27: R and A Values for the Protons of Ethylene I 1/T x 103 Rc{A} 1 XiA{A}{seconds 1) 2 ^AC (seconds 1) 3 4.01 125 1.38 124 4.09 84.1 1.51 82.6 4.18 40.6 2.32 38.3 4.26 24.2 1.43 22.7 4.36 13.2 2.30 10.9 4.45 7.13 2.43 4.70 4.56 5.78 2.67 3.11 4.62 4.38 3.47 0.91 1 Determined by the observation of C nuclei while irradiating A nuclei. 2Determined by the observation of A nuclei after saturation of the same. 3Exchange rate for rotation of ethylene I exchanging nuclei A with nuclei C. Table 5.27 and table 5.28 contain the calculated values of R, R •+• a (Ai) and k for the exchange between protons A and C, and B and D, caused by ethylene rotation at the given temperature. Plots of \n(k/T) versus 1/T for the rotational exchange are shown in figure 5.43 and figure 5.44. The corresponding activation parameters for the processes, calculated from the plots are displayed in table 5.31, calculated values required for the calculation of the activation parameters are listed in table 5.29. Chapter 5. NMR Spectroscopic Studies 141 Figure 5.43: \n(kAC/T) versus 1/T. Chapter 5. NMR Spectroscopic Studies 142 Figure 5.44: \n(kBD/T) versus 1/T. Chapter 5. NMR Spectroscopic Studies 143 Table 5.28: R and A for the Protons Of Ethylene II 1/T x 103 RD {B}{seconds 1) 1 XIB{B](seconds 1) 2 ^DB(seconds *) 3 4.01 31.8 0.848 31.0 4.09 21.3 1.49 19.8 4.18 11.4 1.28 10.1 4.26 7.18 1.84 5.34 4.36 4.13 1.96 2.17 1 Determined by the observation of D nuclei while irradiating B nuclei. 'Determined by the observation of B nuclei after irradiation of the same. 3Exchange rate for rotation of ethylene II, exchanging nuclei B with nuclei D. 5.2.2 Dynamic N M R Computer Fitting Values for the exchange constant were obtained by generating spectra to fit the variable temperature spectra by use of the computer program DNMR3 [155]. If the ethylene ligands are modelled separately a very satisfactory agreement can be obtained by super-imposing the two calculated spectra. A comparison between the experimentally obtained spectra and the computer generated fits is shown in figure 5.45 and figure 5.46. The rates used for the computer generated fits are listed in table 5.30. The parameters used for the fits are listed in Chapter 2. The plot of ln(fc/T) versus 1/T is shown in figure 5.47 (rota-tion of ethylene I) and figure 5.48 (rotation of ethylene II). The corresponding activation parameters are listed in table 5.31. The values calculated for the activation parameters do not show a significant difference between the AH* and AS* values for the two ethylenes in either the data obtained by Chapter 5. NMR Spectroscopic Studies 144 Table 5.29: Refined Data for the Ethylene Rotation Fluxionality 1 1/T x 103 HkAc/T) i n ( w r ) 4.01 -.0702 -2.08 4.09 -1.09 -2.51 4.18 -1.83 -3.16 4.26 -2.34 -3.78 4.36 -3.05 -4.66 4.45 -3.87 • 4.56 -4.26 4.62 -5.47 r -0.991 -0.996 a 29.19 27.74 b -7417.1 7410.3 A' 4.316 x 10"3 4.18 x 10~3 S* 2.20 x 10"4 1.38 x 10"4 Sy/x 5:60 x 10 - 2 0.101 C o 0.416 1.53 96.30 366.85 1The lower section of the table contains the linear regression analyses of the tabulated data. The explanation of these values will be discussed in Section 5.2.3. Chapter 5. NMR Spectroscopic Studies 145 using the saturation transfer experiments or the computer fit data; the AG* values do show a significant difference. The AG* values are inherently more accurate by virtue of the covariance of AH* and AS*, this will be discussed in the next section. The AG* values calculated for the system at 0°C 2 show agreement between the DNMR3 results and those obtained experimentally for the separate ethylene ligand ro-tations. The AG* values also agree in distinguishing between ethylenes with ligand I (containing resonance groups A and C) having a lower energy barrier for rotation than ligand II (containing resonance groups B and D). These values are in agreement with the observation that the resonances due to ethylene II are resolved at higher temperatures than the ethylene I resonances. The values also are consistent with the assignment of ligand I to the site trans to the phenoxide oxygen of higher trans influence and cis to the ketone donor of lower trans influence [154]. 2 The coalescence temperature was estimated to be 0°C from the variable temperature 1E N M R spectra. Chapter 5. NMR Spectroscopic Studies Chapter 5. NMR Spectroscopic Studies 148 Figure 5.47: ]n(kAC/T) versus 1/T from the DNMR3 fits. Chapter 5. NMR Spectroscopic Studies 149 Figure 5.48: ln(kBD/T) versus 1/T from the DNMR3 fits. Chapter 5. NMR Spectroscopic Studies 150 Table 5.30: Data From the DNMR Fits for Ethylene Rotation 1 1/T x 103 k^^iseconds-1) k^MH\seconds-1) H^AC/T) i n ( w r ) 3.85 400 80 0.432 -1.18 3.93 220 45 -0.145 -1.73 4.01 125 30 -0.691 -2.12 4.09 80 20 -1.12 -2.50 4.18 40 10 -1.79 -3.17 4.26 25 5 -2.24 -3.85 4.36 10 3 -3.13 -4.34 4.45 5 1 -3.81 -5.41 4.56 2 0 -4.70 4.62r 0 0 r -0.998 -0.993 a 26.53 24.98 b -6780.6 -6765.0 X 4.19 x 10-3 4.14 x 10-3 Sx 2.53 x 10"4 2.10 x 10~4 SY/X 0.102 0.180 0.599 1.34 142.6 324.3 JThe lower section of the table contains the linear regression analyses of the tabulated data. The explanation and significance of these values will be discussed in Section 5.2.3. Chapter 5. NMR Spectroscopic Studies 151 Table 5.31: Activation Parameters Calculated for Ethylene Rotation and Estimated Standard Deviations (in brackets) AH* (Kcal/mol) AS*(cal/mol(K)) AG* (Kcal/mol) Experimental: 1 Ethylene I 14.7(0.2) 10.8(0.8) 11.8(0.03) Ethylene II 14.7(0.7) 7.91(3.04) 12.5(0.1) DNMR3: Ethylene I 13.5(0.3) 5.50(1.19) 12.0(0.04) Ethylene II 13.4(0.6) 2.42(2.66) 12.7(0.08) Determined from the saturation transfer experiments. Chapter 5. NMR Spectroscopic Studies 152 5.2.3 Error Calculations The errors in AH* and AS* are estimated by the calculation of the standard deviation in the slope and F-intercept in the plot of \n(k/T) versus 1/T from which the activation parameters were derived. These calculations are described by Kleinbaum and Kupper [156] and are summarized below (equation 5.21 to equation 5.24). The standard deviation for AH* equals Ofe and the standard deviation for AS* equals <ra. The standard deviation in AG* is relatively lower than that of its component AH* and AS* values due to the covariance between these components [121]. The standard deviation in AG*T is calculated by equation 5.25, the error lost in AG*T due to the covariance of AH* and AS* is reflected in the 2TaAHtcr&s< term. 1 E(F - Y)2 (5.21) n - 2 < ? 2 Y/X — Variance of Y given X. Line form : Y = a -f bX Y = The value of Y predicted by the least squares determined line for a given A'. n = The number of A', Y data points. (5.22) Sx>/n - 1 (5.23) sx = - ± - m t - x ) 2 n — 1 (5.24) Chapter 5. NMR Spectroscopic Studies 153 X = The mean of all A r,. Xi = Individual X values. ( ^ ) 2 = KHO 2 + (TaASi)2 - (2T crAHtcrASt) (5.25) 5.2.4 Ethylene Exchange Irradiation of one of the proton resonances on one of the ethylene ligands causes the reso-nances of the other ligands to decrease in intensity. This effect is displayed in figure 5.49 where irradiation of the resonance A at 4.1 ppm, due to the outer protons on ligand I, causes a decrease in the C (3.7 ppm) resonance (due to the ethylene rotation) but also causes a decrease in resonances B (2.4 ppm) and D (2.2 ppm). The decrease in B and D upon the irradiation of A is indicative of exchange between ethylene ligands. The amount of this decrease can be used to calculate the exchange rate. The necessary data can be obtained from experiment 1 which was used for the determination of R for ethy-lene rotation. Here NOE is no longer a factor as the protons on separate ethylene ligands are not coupled to each other. Again R is calculated from the slope of the ln(/t — loo) versus time plot. In this experiment the protons in an ethylene ligand are observed at various evolution times after a set of protons on the other ethylene ligand is decoupled. In the absence of NOE it is possible to obtain the ratio o~/R from the steady state mag-netization (Id + I0j((Tij/Ri); I0i — I0j). The rate of fluxionahty of the two ligands is not the same; irradiation of a set of protons on ethylene II causes a greater decrease in the intensity of the ethylene I resonances than vice versa. This effect can be seen in the variable temperature decoupled spectra. In figure 5.50, saturating the B resonance has a much greater effect on resonances A and C than the effect saturating the A resonance has on B and D. Saturation of the B resonance effects the complete saturation of A and Chapter 5. NMR Spectroscopic Studies 154 Figure 5.49: The effects of irradiation of A on resonances B and D (top) and the effects of irradiation of B on resonances A and C (bottom) at -38.5°C. C resonances at a temperature of —28.5°C. A temperature of higher than —23.5°C is required for the irradiation of A to completely saturate B and D. Intramolecular exchange by a 'Berry pseudorotation', suggested by Kreiter et al. for bis(7/2-olefin)(2,4-pentanedionato-0,0 ' )rhodium(I) derivatives, discussed in the intro-duction of this chapter, does not seem to be occurring here. The difference in the values of kg{Ay (rate of exchange calculated for B during the irradiation of A) and kA{B} (ob-tained for calculation of k exchange during irradiation of B) indicate that the exchange between ethylenes is not intramolecular; otherwise the observed rates would be equal. An intermolecular process involving the collision of two molecules of complex and exchange of ethylene also does not account for the disparity in the observed exchange rates as, again, they are expected to be equal. Further evidence negating this type of exchange is garnered from the study of the COD analogue of 15, complex 18, described next. Chapter 5. NMR Spectroscopic Studies 15 CL Chapter 5. NMR Spectroscopic Studies 156 J H N M R of (7/ 4-l,5-Cyclooctadiene)(2-acetylphenoxy-0,0 ')rhodium(I), 18 The COD analogue of 15 is an ideal model for the study of the possibility of intramolec-ular exchange in complexes of this type. It is expected that if a 'Berry pseudorotation' process is occurring in 15 it should also occur in 18. The complete : H NMR spectrum of 18 at room temperature is shown in figure 5.51 and has been assigned in Chapter 3. One of the advantages of using complex 18 is that it has only two, well separated, resonances for the olefinic protons in the *H NMR spectrum. Another feature of this system is that the olefinic protons are not coupled to each other. This feature should allow easy study of exchange by using the spin saturation transfer method since NOE should not be a problem. Variable temperature 1 H NMR spectra of 18 run at temperatures ranging from H-30°C to —30°C show no changes associated with an exchanging system (figure 5.52). The loss of the fine structure in the olefinic resonances at lower temperatures is attributed to the dependence of spectral quality on temperature. The position of the olefinic proton resonances in the spectrum of 18 (4.13 ppm, 4.43 ppm) supports the assignment of the ethylene resonances in the low temperature *H NMR spectrum of 15. The olefinic protons in 18 are structurally analogous to the 'outer' protons in 15, as COD has no 'inner' protons, and he in the same region in the NMR spectrum (the 'outer' ethylene protons appear at 3.5 ppm and 4.2 ppm in 15). The shielding effect of the COD backbone upon the olefinic protons in 18 is expected to be negligible. Observation of the cyclooctadiene analogue of 15 shows the absence of exchange be-tween the olefinic protons. Similarly, no intramolecular exchange should occur between the ethylene ligands of 15. The ethylene exchange can then be addressed as an inter-molecular process. The mechanism of the ethylene exchange appears to involve free ethylene and the 16 Chapter 5. NMR Spectroscopic Studies 157 i ' ' i | i i ' ' i i I I ' | ' ' i ' i i I I i | i i I I i ' i i i I i i i i i i i i i I I I I I i I I I I I i i I I | i i i i I i i i 7.0 6.0 5.0 4.0 3.0 2.0 1.0 PPM Figure 5.51: The 1 H NMR spectrum of 18 at room temperature (assignments can be found in table 3.9). Chapter 5. NMR Spectroscopic Studies 158 + 3 0 . 7 ° C •+17.4°C + 9 . 5 ° C -5.5°C -20.2°C A I 1 -34.7°C " I "I I I | I I H | I I I I | i I I ' • • • 5A 43 AA 3.9 3.4 ppm Figure 5.52: The variable temperature 1 H NMR spectrum of 18. Chapter 5. NMR Spectroscopic Studies 159 electron bis(ethene)rhodium(I) complex. A schematic representation of the proposed re-action mechanism is shown in figure 5.53. The first step in the reaction is the dissociation of ethylene, this process must be reversible, since no decomposMon or change in the 1 H NMR spectrum is observed over long periods. The presence of any significant amount of ethylene in solution would give rise to a peak in the NMR spectrum unless if it were un-dergoing rapid exchange. If free ethylene were present and rapid exchange were occurring the resonances due to the protons with which it was exchanging (i.e. the ethylene ligand protons) would be shifted downfield towards the position of free ethylene. The chemical shift for the ethylene ligand resonances would be expected to be temperature dependent, also owing to this effect. Neither of these effects is observable. This behavior indicates that the equilibrium for the dissociative reaction favours the IB electron complex. The second step of the overall process involves the reaction between ethylene and either the 16 electron complex or the 14 electron intermediate. If the ethylene associates to a vacant site on a 14 electron intermediate the association will only be observable if it involves ethylene site exchange (i.e. ethylene going from siteT to site II or from site II to site I). This selective measurement results from the experimental method in which spin saturation transfer can only be attributed to exchange with the second type of ethylene ligand (the one being saturated). The favoured reaction is between ethylene and the 16 electron complex to give a five coordinate intermediate. Such an intermediate has been postulated for the exchange reaction between gaseous ethylene and bis(772-ethene)(2,4-pentanedionato-0,0')rhodium(I) [55]. The exchange can then be explained by dissoci-ation of one of the ethylene ligands from the five coordinate complex and migration of the 'new' ethylene into the vacant site. The rate law for this substitution reaction con-sists of two terms (equation 5.26) where [complex00*] is the concentration of the observed complex, [complex '] is the initial concentration of complex (or the concentration of rhodium) and [ethylene] is the concentration of ethylene in solution. Since the [ethylene] Chapter 5. NMR Spectroscopic Studies 160 << [complex] the second term is ignored and the rate law becomes equation 5.27. dlcomplex1*'} = k2[complexl] 4- ki[complexl}[ethylene] (5.26) dt d[complex°b''] = k2[complex] (5.27) dt The calculated values of R, <r/R for the ethylene ligand protons and k are shown in table 5.32. The rate constant k2 is equal to A;/[complex] ([complex] = 0.016 M). The data required for the ln(fc/T) versus 1/T plots and error data are shown in table 5.33. It was not possible to acquire data by observation of the D protons as too much overlap with from the slope and Y intercept from these plots. Error calculations were also performed as previously described; a single error calculation is performed for observation of A and C exchange, as the intermolecular exchange does not discriminate between ethylene sites on individual ethylene ligands. Activation parameters are shown in table 5.34. The results of the study of the intermolecular exchange are consistent with the results from the study of ethylene rotation. In both cases ethylene I is undergoing more rapid exchange with lower activation barriers to the motion than ethylene II. The ethylene trans to the c-bound oxygen has the lower activation barrier for the dissociation reaction which is expected, and also confirms the initial ethylene ligand assignment. the solvent peak occurs at these temperatures. Plots of ]n(k/T) are shown in figure 5.54 (measured by the observation of the protons of ethylene I) and figure 5.55 (measured by observing the protons of ethylene II). Again the activation parameters were calculated Chapter 5. NMR Spectroscopic Studies Figure 5.53: Proposed reaction mechanism for intramolecular ethylene exchange. Chapter 5. NMR Spectroscopic Studies 162 Figure 5.54: Plot of \n(k2/T) versus 1/T for exchange of ethylene I. Chapter 5. NMR Spectroscopic Studies 163 1.5-1 Figure 5.55: Plot of \n(k2/T) versus 1/T for exchange of ethylene II. Chapter 5. NMR Spectroscopic Studies 164 Table 5.32: R and a/R for Intermolecular Ethylene Exchange 1/T x 103 RB{A} RA{B} Rc{B} (<r/R)({B}) 3.85 14.1 -1 4.01 4.48 27.1 26.8 -0.698 -1 4.09 3.67 16.0 14.8 -0.545 -1 4.18 3.29 7.15 6.73 -0.263 . -0.846 4.26 5.59 4.46 -0.673 4.36 3.11 3.75 -0.503 Chapter 5. NMR Spectroscopic Studies 165 Table 5.33: Calculated Data for Ethylene Dissociation 1/T x 10 3 ln(*ic {B } / r ) 3.85 1.22 4.01 -0.241 1.91 1.91 4.09 -0.671 1.41 1.33 4.18 -1.49 0.457 0.397 4.26 1.10 x 1 0 " 3 -0.233 4.36 -0.849 -0.656 r -0.997 -0.993 a 32.28 33.25 b -8078 -7820 X 4.0325 x 10~ 4 4.18 x 1 0 - 3 Sx 1.401 x 1 0 - 3 1.298 x 10 " 4 SA'/V 0.1136 0.1285 Va 1.89 1.38 <7b 468 330.1 Table 5.34: Calculated Activation Parameters for Dissociaton Ethylene I (St. Dev.) Ethylene II (St. Dev.) AH* AS* 15.5(0.7)(Kcal/mol) 18.9(2.7)(cal/mol K) 9.83(0.17)(Kcal/mol) 16.1(0.9)(Kcal/mol) 16.9(3.8)(cal/mol K) 11.0(0.2)(Kcal/mol) Chapter 5. NMR Spectroscopic Studies 166 5.3 Complex 15 in Methanol Unlike the spectrum of 15 in toluene- d8 described above, the variable temperature spec-tra of 15 in methanol-^ are complex (figure 5.56). As the temperature is lowered res-onances corresponding to one ethylene resolve quickly. Upon further lowering of the temperature two sets of resonances each corresponding to one ethylene are resolved. A tentative explanation which is consistent with the observed NMR spectra has been devel-oped to account for these results. The J H NMR can be explained by the presence of two species: a (methanol)( 772-ethene)rhodium(I) complex 44 in which an ethylene has been replaced, and a five coordinate bis(7/2-ethene)(methanol)rhodium(I) complex 45. Com-plex 44 would account for the ethylene ligand which is resolved at —33°C, since it would be expected that the more labile ethylene would be displaced, placing the methanol trans to the phenoxide oxygen, leaving the ethylene which has been shown to have a higher barrier to rotation. The rest of the spectrum may be explained by the presence of the five coordinate complex 45. This species has ethylene ligands with the following characteristics: the ethylene ligands are in similar magnetic fields, evident by their chemical shifts, and they appear to have the same coalescence temperature. The first observation is in accord 44 45 Chapter 5. NMR Spectroscopic Studies 167 4.0 3.0 2.0 ppm Figure 5.56: Ethylene region of 15 in methanol-d4. Chapter 5. NMR Spectroscopic Studies 168 with the possible geometries (square pyramid or trigonal bipyramid). In either geometry an ethylene ligand would not be directly trans to the phenoxide oxygen; hence a strong trans effect would not be expected. The new structure would also move the ethylene, originally cis to the ketone, out of the ketone deshielding cone. A combination of these effects can account for the similar chemical shifts for the ethylene resonances. The similar coalescence temperatures for the two ethylene resonances can be explained by a Berry pseudorotation. This motion would exchange ethylenes at an equivalent rate and is possible for 5-coordinate complexes of ds metals as discussed in the introduction of this chapter. Exchange between the free ethylene and the coordinate ethylene is likely as a reso-nance due to free ethylene is evident in the spectrum at 5.2 ppm. A sample of 15 in toluene- ds with ethylene gas (60 mm pressure) was prepared to explore the effects of the free ethylene (complex : free ethylene in solution % 2 : 1 by integration). The free ethylene made little effect on the overall appearance of the 1 H NMR spectrum aside from the resonance at 5.2 ppm which is only evident at low temperatures. The presence of the ethylene gas did not give rise to resonances which would be associated with a five coordinate species. Irradiation of the ethylene peak caused a measurable decrease in the complexed ethylene resonances. This effect was observed at a variety of temperatures; a greater decrease in intensity was noted in the resonances of the ethylene I protons than in those of the ethylene II protons during saturation of the free ethylene signal. Again, it is evident that ethylene I has a higher exchange rate than ethylene II. A larger saturation transfer effect on the ethylene I resonance is consistent with the measured values of A G * for the ethylene exchange process. Chapter 6 Summary The bis(olefm)rhodium(I) complexes studied here were found to be interesting in a num-ber of respects. A synthetic route was developed to prepare the complexes in high yield under anhydrous conditions. The complexes were characterized by elemental analysis, 1 H NMR spectroscopy, and mass spectrometry. All of the rhodium complexes employed gave rise to parent molecular ions in the mass spectrum. Crystallographic studies confirmed the structure of the studied complexes and indicated some interesting features. The complexes were found to be precursors for active homogeneous olefin hydro-genation catalysts and are the first reported examples of catalysts using this type of rhodium/ligand system. The catalysts show selectivity for the hydrogenation of sterically unhindered double bonds in the presence of hindered double bonds and are effective in the presence of aromatic, alcohol and carboxylic acid functional groups on the olefinic sub-strate. The catalysts can be prepared in situ by the addition of the bis(ethene)rhodium(I) chloride dimer to a solution of the sodium salt of the appropriate ligand in the reaction vessel prior to the catalytic reaction. The rates of the catalytic hydrogenation reaction using these rhodium(I) complexes are comparable to those of existing rhodium(I) catalyst systems for the same substrates. The complexes were also found to be useful catalysts for the homogeneous hydrosi-lylation of ketones effecting moderate conversions to the corresponding silyl ether under mild reaction conditions. Attempts to prepare an asymmetric catalyst using this type of system were not entirely successful; optical yields did not exceed 9%e.e.. 169 Chapter 6. Summary 170 The bis(ethene)rhodium(I) complexes were found to exhibit fluxional behavior in their X H NMR spectra associated with the motion of the ethylene ligands. The fluxionality of one of the catalyst precursors, 15, was studied in detail. The ethylene fluxionality originates from two separate processes, namely, ethylene rotation and ethylene exchange. These processes were studied separately by the use of specialized NMR techniques. The rotational exchange was also modeled using the DNMR3 computer fitting program. Ex-change rates and activation parameters were calculated for the two exchange processes. Excellent agreement was obtained between the AG* values calculated from spin satura-tion NMR experiments and those obtained by the use of the computer fitting program for the ethylene rotation process. The AG* values for the intermolecular exchange'showed significant differences between the ethylene ligands revealing that the two ethylenes in bis(ethene)rhodium(I) complexes can behave independantly. This result opposes the theory that the ethylene ligand motion is coupled by a gear type of mechanism. Bibliography [1] a) Hughes, R.P. In Comprehensive Organometallic Chemistry; Wilkinson, G.; Stone, F.G.A.; Abel, E., Eds.; Permagon: New York, 1982; Volume 5, Chapter 35; b) Ryan, T.A. Coord. Chem. Rev. 1984, 57, 75; c) Cole-Hamilton, D.J. Coord. Chem. Rev. 1981, 35, 113. [2] a) James, B.R. In Comprehensive Organometallic Chemistry; Wilkinson, G.; Stone, F.G.A.; Abel,' E.W., Eds.; Permagon: New York, 1982; Volume 8, 285; b) Jardine, F.H. Prog. Inorg. Chem. 1981, 28, 63; c) James, B.R. Homogeneous Hydrogenation; Wiley-Interscience: New York, 1973. [3] Parshall, G.W. Homogeneous Catalysis, the Applications and Chemistry of Catal-ysis by Soluble Transition Metal Complexes; Wiley-Interscience: New York, 1980. [4] Collman, J.P.; Hegedus, L.S.; Norton, J.R.; Finke, R.G. Principles and Applications of Organotransition Metal Chemistry; University Science Books: CA, 1987. [5] Sanchez-Delgado, R.A.; DeOchoa, D.L. J. Mol. Catal. 1979, 6, 303. [6] Ojima, I.; Hirai, K. In Asymmetric Synthesis; Morrison, J.D., Ed.; Academic: New York, 1985; Volume 5, Chapter 4. [7] Garrou, P.E. Chem. Rev. 1985, 171. [8] Tolman, C A . Chem. Rev. 1977, 77, 313. [9] McAuliffe, C A . Transition Metal Complexes of Phosphorus, Arsenic and Antimony Ligands; Wiley: New York, 1973. 171 Bibliography 172 [10] McAuhfTe, C.A.; Levanson, W. Phosphine, Arsine, and Stibine Complexes of Tran-sition Elements; Elsevier: New York, 1979. [11] Zassinovich, G.; Camus, A.; Mestroni, G. Inorg. Nucl. Chem. Letters 1976, 12, 865. [12] a) Zassinovich, G.; Del Bianco, C ; Mestroni, G. J. Organomet. Chem. 1981, 222, 323; b) Uson, R.; Oro, L.A.; Sariego, R.; Valderrama, M.; Rebulhda, C. J. Organomet. Chem. 1980, 197, 87. [13] Birch, A.D.; Williamson, D.H. Org. Reac. 1974, 24, 1. [14] Jardine, I.; McQuillin, F.F. Chem. Comm. 1969,477. [15] a) Zassonovich, G.; Mestroni, G. J. Mol. Catal. 1987, J^2, 81; b) Spogliarich, R.; Kasper, J.; Graziani, M. J. Catal. 1985, 94, 292; c) Beaupere, D.; Nadjo, L.; Uzan, R.; Bauer, P. J. Mol. Catal. 1983, 20 (2), 195; d) Grigg, R.; Mitchell, R.B.; Tongpenyai, N. Synthesis 1981, 442. [16] a) Pittman, C U . Jr. In Comprehensive Organometallic Chemistry; Wilkinson, G.; Stone, F.G.A.; Abel, E.W., Eds.; Permagon: New York, 1982; Volume 8, Chapter 55; b) Dumont, W; Poulin, J .C ; Dang, T.P.; Kagan, H. J. Am. Chem. Soc. 1973, 95, 8295; c) Collman, J.P.; Hegedus, L.J.; Cooke, M.P.; Norton, J.R.; Dolcetti, G.; Marquardt, D.N. J. Am. Chem. Soc. 1972 , 94, 1789; d) Capka, M.; Svoboda, P.; Creny, M.; Hetflejs, J. Tetrahedron Lett. 1971, 4787. [17] Fryzuk, M.D.; Bosnich, B. J. Am. Chem. Soc. 1977, 99, 6262. [18] Fryzuk, M.D.; Bosnich, B. J. Am. Chem. Soc. 1978, 100, 5491. Bibliography 173 [19] a) Knowles, W.A. Acc. Chem. Res. 1983, 16, 106; b) Brown, J.M.; Chaloner, P.A. J. Chem. Soc, Chem. Commun. 1980, 344. [20] Vineyard, B.D.; Knowles, W.S.; Sabachy, M.J.; Bachman, G.L.; Weinkauff, D.J. J. Amer. Chem. Soc. 1977, 99, 5946. [21] Hayashi, T.; Yamamoto, K.; Kumada, M. Tetrahedron Lett. 1974, 4405. [22] a) Spencer, A. In Comprehensive Coordination Chemistry; Wilkinson, G.; Gillard, R.D.; McCleverty, J.A., Eds.; Permagon: New York, 1987; Volume 6, Chapter 31, 250; b) James, B.R.; Joshi, A.; Kvintovics, P.; Morris, R.M.; Thorburn, I.S. In Catalysis of Organic Reactions; Blackburn, D.W.; Dekker, M., Eds.; In press. [23] Kagan, H.B. In Comprehensive Organometallic Chemistry; Wilkinson, G.; Stone, F.G.A.; Abel, E.W., Eds.; Permagon: New York, 1982; Volume 8, Chapter 53, 463. [24] Ojima, I.; Yamamoto, K.; Kumada, M. In Aspects of Homogeneous Catalysis; Ugo, R., Ed.; Reidel: Dordrecht, 1977; Volume 3, Chapter 3. [25] Zeise, W.C. Pogg. Ann. Phys. Chem. 1827, 9, 632. [26] Kauffman, G.B. In Comprehensive Coordination Chemistry; Wilkinson, G.; Gillard, R.O.; McCleverty, J.A., Eds.; Permagon: New York, 1987; Volume 1, Chapter 1, 1. [27] Ittel, S.D.; Ibers, J.A. Adv. Organomet. Chem. 1976, 14, 33. [28] Mingos, D.M.P. In Comprehensive Organometallic Chemistry; Wilkinson, G.; Stone, F.G.A.; Abel, E.W., Eds.; Permagon: New York, 1982; Volume 3, Chapter 19, 1. [29] Dewar, M.J.S. Bull. Soc. Chim. France 1951, 18, C79. Bibliography 174 [30] Chatt, J.; Duncanson, L.A. J. -Chem. Soc. 1953, 2939 [31] Cramer, R. J. Am. Chem. Soc. 1967, 89, 4621. [32] a) Mann, B.E. In Comprehensive Organometallic Chemistry; Wilkinson, G.; Stone, F.G.A.; Abel, E.W., Eds.; Permagon: New York, 1982; Volume 3, Chapter 20, 89; b) Greaves, E.O.; Lock, C.J.L.; Maiilis, P.M. Can. J. Chem. 1968, 69, 799. [33] Hartley, F.R. Chem. Rev. 1968, 69, 799. [34] Quin, H.W.; Tsai, J.H. Adv. Inorg. Radiochem. 1969, 12, 217. [35] Hartley, F.R. Angew. Chem. 1972, 11, 596. [36] Mingos, D.M.P. Adv. Organomet. Chem., 1977, 15, 1. [37] Schrock, R.R.; Osbbrn, J.A. J. Am. Chem. Soc. 1971, 93, 2397. [38] Ojima, I.; Kogure, T.; Kumagai, M.; Horiuchi, S.; Sato, T. J. Organomet. Chem. 1976, 122, 83. [39] Hayashi, T.; Yamamoto, K.; Kasuga, K; Omizu, H.; Kumada, M. J. Organomet. Chem. 1976, 113, 127. [40] Shrock, R.R.; Osborn, J.A. J. Amer. Chem. Soc. 1971, 94, 2397. [41] Crabtree, R.H.; Gautier, A.; Giordano, G.; Khan, T. J. Organometal. Chem. 1977, 141, 113. [42] a) Combes, A. Ann. Chim. 1887, 12, 199; b) Combes, A. Compt. Rend. 1887, 105, 869; c) Combes, A. Compt. Rend. 1889, 108, 415. Bibliography 175 [43] a) Siedle, A.R. In Comprehensive Coordination Chemistry; Wilkinson, G.; Gillard, R.D.; McCleverty, J.A., Eds.; Permagon: New York, 1987; Volume 6, Chapter 61, 250; b) Joshi, K.C. Pathak, V.N. Coord. Chem. Rev. 1977 , 22, 37; c) Bock, B.; Flatau, K.; Junge, H.; Kuhr, M.; Musso, H. Angew. Chem., Int. Ed. Engl. 1971, 10, 225; d) Graddon, D.P. Coord. Chem. Rev. 1969, 4, 1; e) Fackler, J.P.Jr. Prog. Inorg. Chem. 1966, 7, 362. [44] Mehrotra, R.C; Bohra, R.; Gaur, D.P. Metal B-Diketonates and Allied Derivatives; Academic Press: New York, 1978. [45] Crabtree, A.H.; Lavin, M.; Bonneviot, L. J. Am. Chem. Soc. 1986, 108, 4032. [46] Whitesides, C M . ; Lewis, D.W. J. Am. Chem. Soc. 1970, 92, 6979. [47] Vrieze, K.; Van Leeuwen, P.W.N.M. Prog. Inorg. Chem. 1974, 14, 1. [48] Marquarding, D.; Klusacek, H.; Gokel, C ; Hoffman, P.; Ugi, I. J. Am. Chem. Soc. 1970, 92, 5389. [49] Grundy, J.; James, B.G.; Pattenden, G. Tetrahedron Lett. 1972, 9, 757. [50] Alvarez, F.S.; Watt, A.N. J. Org. Chem. 1968, 33, No. 5, 2143 [51] Durst, H.D.; Gokel, G.W. Experimental Organic Chemistry; McGraw-Hill: Toronto, 1980. [52] Efimovsky, 0. J. Recherches Center Natl. Recherche Sci. 1959 47, 147. [chem. abstr. 4742f (1962)]. [53] Noggle, J.H.; Shirmer, R.E. The Nuclear Overhauser Effect, Chemical Applications; Academic: New York, 1971. Bibliography 176 [54] Cramer, R.; Kline, J.B.; Roberts, J.D. J. Am. Chem. Soc. 1969, 91, 2519. [55] Cramer, R. J. Am. Chem. Soc. 1964, 86, 217. [56] Mestroni, G.; Zassonovich, G.; Camus, A.; J. Organomet. Chem. 1977, 140, 63. [57] Brunner, H.; Beier, P.; Riepl, G.; Bernal, I.; Reisner, G.M.; Benn, R.; Rufinska, A. Organometallics 1985, 4, 1732. [58] King, R.B.; King, A.D.; Igbal, M.Z. J. Am. Chem. Soc. 1979, 101, 4893. [59] Van Eldik, R.; Aygen, S.; Kelm, H. Trans. Met. Chem. 1985, 10, 167. [60] Vidal, J.L.; Walker, W.E.; Pruett, R.L.; Shoening, R.C; Fiato, R.A. Fundam. Res. Homogeneous Catal. 1979, 3, 499. [61] Vidal, J.L.; Walker, W.E. Inorg. Chem. 1980, 19, 896. [62] Ryan, T.A. Coord. Chem. Rev. 1982, 41, 251. [63] Trezciak, A.M.; Ziolkowski, J.J.; Aygen, S.; Van Eldik, R. J. Mol. Catal. 1986, 34, 337. [64] Janecko, H.; Trzeciak, A.M.; Kiolkowski, J.J. J. Mol. Catal. 1984, 26, 355. [65] Pieta, D.; Trzeciak, A.M.; Kiolkowski, J.J. J. Mol. Catal. 1983, 18, 193. [66] Whitmore, B.C.; Eisenberg, R. J. Am. Chem. Soc. 1984, 106, 3225. [67] Sacconi, L. Coord. Chem. Rev. 1966, 1, 192. [68] Calderazzo, F.; Floriani, C ; Henzi, R.; L'Eplattenier, F. J. Chem. Soc. (A) 1969, 1378. [69] Mague, J.T.; Nutt, M.O. J. Organomet. Chem. 1979, 166, 63. Bibliography 177 [70] Cozens, R.J.; Murray, K.S.; West, B.O. Aust. J. Chem. 1970, 23, 683. [71] Cozens, R.J.; Murray, K.S.; West, B.O. J. Organomet. Chem. 1971, 27, 399. [72] Elsevier's Encyclopaedia of Organic Chemistry; Josephy, E.; Radt, F., Eds.; Else-vier: New York, 1948; Volume 12A, Series III. [73] Cain, C.E.; Mashburn, A.T.Jr.; Hauser, C.R. J. Org. Chem. 1961, 26, 1030. [74] Gordon, A.J.; Ford, R.A. The Chemist's Companion; Wiley and Sons: New York, 1972. [75] Anton, D.A.; Crabtree, R.H. Organometallics 1983, 2, 855. [76] Osborn, J.A.; Jardine, F.H.; Young, J.F.; Wilkinson, G. J. Chem. Soc. (A) 1966, 1711. [77] Birch, A.J.; Walker, K.A.M. J. Mol. Catal. 1966, 1894. [78] Zuber, M.; Szuberla, W.A.; Pruchnik, F.P. J. Mol. Catal. 1986, 38, 309. [79] Hon, P.K.; Belford, R.L.; Pfluger, C E . J.Chem.. Phys. 1965, 43, 1323. [80] Hon, P.K.; Pfluger, C.E.; Belford, R.L. Inorg. Chem. 1966, 5, 516. [81] Hon, P.K.; Pfluger, C.E.; Belford, R.L. Inorg. Chem. 1967, 6, 730. [82] Belford, R.L.; Chasteen, N.D.; Hitchman, M.A.; Hon, P.K.; Pfluger, C.E.; Paul, I.C Inorg. Chem. 1969, 8, 1312. [83] Cotton, F.A.; Legzdins, P. Inorg. Chem. 1968, 7, 1777. [84] Zalkin, A.; Templeton, D.H.; Karraker, D. Inorg. Chem. 1969, 8, 2680. Bibliography 178 [85] Pantini, G.; Racanelli, P.; Immirzi, A.; Porri, L. J. Organomet. Chem. 1971, 33, C17. [86] Ewings, P.F.R.; Harrison, P.G.; King, T.J. J. Chem. Soc., Dalton Trans. 1975, 1455. [87] Bancroft, M.; Davies, B.W.; Payne, N.C.; Sham, T.K. J. Chem. Soc, Dalton Trans. 1975, 973. [88] Leipoldt, J.G.; Bok, L.D.C.; Van Vollenhoven, J.S.; Pieterse, A.I. J. Inorg. Nucl. Chem. 1978, 40, 61. [89] Leipoldt, J.G.; Basson, S.S.; Lamprecht, G.J.; Bok, L.D.C.; Schlebusch, J.J.J. Inorg. Chim. Acta. 1980, 40, 43. [90] Evans, J.A.; Russel, D.R. J. Chem. Soc. Chem. Commun. 1971, 197. [91] CRC Handbook of Chemistry and Physics; Weast, R.C.; Astle, M.J., Eds.; CRC: Florida, 1980. [92] Rettig, S.J.; Trotter, J. Can. J. Chem. 1976 , 54, 1168. [93] Rosenblum, M. Chemistry of the Iron Group Metallocenes; John Wiley: New York, 1965. [94] a) Halpern, J.; Okamoto, T.; Zakhariev, A. J. Mol. Catal. 1976, 2, 65; b) Dawans, F.; Morel, D. J. Mol. Catal. 1977, 403. [95] a) Chan, A.S.C.; Halpern, J. J. Amer. Chem. Soc. 1980, 102, 838; b) Halpern, J.; Riley, D.P.; Chan, A.S.C.; Pluth, J.J. J. Amer. Chem. Soc. 1977, 99, 8055; c) Halpern, J. Science 1982, 217, 401. Bibliography 179 [96] Van Uitert, L.G.; Fernelius, W.C.; Douglas, B.E. J. Am. Chem. Soc. 1953, 75, 2736. [97] Van Uitert, L.G.; Fernelius, W.C.; Douglas, B.E. J. Am. Chem. Soc. 1953, 75, 457. [98] Watts, V.S.; Goldstein, J.H. In The Chemistry of Alkenes; Zabicky, J., Ed.; Inter-science: London, 1970; Volume 2, Chapter 1. [99] Crabtree, R. Acc. Chem. Res. 1979, 12, 331. [100] Khan, M.M.T.; Martell, A.E. Homogeneous Catalysis by Metal Complexes; Aca-demic Press: New York, 1974; Volume 1, Chapter 1. [101] a) Meals, R.N. Pure Appl. Chem. 1966, IS, 141; b) Seyferth, D.; Weinstein, R.M. J. Am. Chem. Soc. 1982, 104, 5534; c) Reich, H.J.; Kelly, M.J. J. Am. Chem. Soc. 1982, 104, 119-[102] a) Speier, J.L. Adv. Organomet. Chem. 1978, 17, 407; b) Cundy, C.S.; Kingston, B.M.; Lamppert, M.T. Adv. Organomet. Chem. 1973, 11, 253. [103] a) Magnus, P.D.; Sarker, T.; Djuric, S. In Comprehensive Organometallic Chem-istry; Wilkinson, G.; Stone, F.G.A.; Abel, E.W., Eds.; Permagon: New York, 1982; Volume 7, Chapter 48, 515; b) Brownbridge, P. Synthesis 1983, 1, 1; c) Colvin E.W. Silicon in Organic Synthesis; Butterworth and Co.: London, 1981; d) Weber, W.P. Silicon Reagents for Organic Synthesis; Springer-Verlag: Berlin, 1983. [104] Brunner, H. Angew. Chem. 1983, 22, 897. [105] a) Ojima, I.; Nihonyanagi, M.; Nagai, Y. J. Chem. Soc, Chem. Commun. 1972, Bibliography 180 938; b) Ojima, I.; Nihonyanagi, M.; Kogure, T.; Kumagai, M.; Horiuchi, S.; Nakat-sugawa, K.; Nagai, Y. J. Organomet. Chem. 1975, 94, 449. [106] a) Cullen, W.R.; Evans, S.V.; Han, N.F.; Trotter, J. Inorg. Chem. 1987, 26, 514; b) Brunner, H.; Reifer, B.; Riepl, G. Chem. Ber. 1984, 117, 710; c) ibid, 1330; d) Yamamoto, K.; Hayashi, T.; Kumada, M. J. Organomet. Chem. 1971, 28, C37; e) Speier, J.L. Adv. Organomet. Chem. 1978, 17, 407. [107] a) Takahashi, S.; Shibano, T.; Hagihara, S. J. Chem. Soc. Chem. Commun. 1969, 169; b) Tsuji, J.; Hara, M.; Ohno, K. Tetrahedron 1974, 30, 2143. [108] Yamamoto, K.; Hayashi, T.; Kumada, M. J. Organomet. Chem. 1972, 46, C65. [109] Chan, A.S.C; Landis, C.R.; Shieh, H.S., personal communication. [110] Ojima, I.; Kogure, T.; Kumagai, M.; Horiuchi, S.; Sato, T. J. Organomet. Chem. 1976, 122, 83. [Ill] Dumont, W.; Poulin, J .C; Dang, T.P.; Kagan, H.B. J. Amer. Chem. Soc. 1973, 95, 8295. [112] Johnson, T.H.; Klein, K.C.; Thomen, S. J. Mol. Catal. 1981, 12, 37. [113] Hayashi, T.; Kumada, M.; Yoshida, K. In Fundamental Research in Homogeneous Catalysis; Ishii, Y.; Tsutsui, M., Eds.; Plenum: New York, 1978; Volume 2, 159. [114] Hayashi, T.; Yamamoto, K.; Kumada, M. Tetrahedron Lett. 1974, 49, 4405. [115] a) Brunner, H.; Riepl, G. Angew. Chem. Int. Ed. Eng. 1982, 21, No.5, 377; b) Brunner, H.; Becker, R.; Riepl, G. Organometallics 1984, 3, 1354. [116] Brunner, H.; Knott, A.; Benn, R.; Rufinska, A. J. Organomet. Chem. 1985, 295, 211. Bibliography 181 [117] a) Hayashi, T.; Kumada, M. Acc. Chem. Res. 1982, 15, 395; b) Appleton, T.G.; Cullen, W.R.; Evans, S.V.; Kim, T.J.; Trotter, J. J. Organomet. Chem. 1985, 279, 5. [118] Marquarding, D.; Klusacek, H.; Gokel, G.; Hoffman, P.; Ugi, I. J. Am. Chem. Soc. 1970, 92, 5389. [119] a) Cullen, W.R.; Woolens, J.D. Coord. Chem. Rev. 1981, 39, 1; b) Butler, I.R.; Cullen, W.R; Kim, T.J. Synth. Reac. Inorg. Met.-Org. Chem. 1985, 15, 109. [120] a) Reeve, W.; Group, E.F.Jr. J. Org. Chem. 1967, 32, 122; b) Moise, C ; Tirouflet, J.; Singer, H. Bull. Soc. Chim, Fr. 1969, 4, 1182. [121] Sandstrom, J. Dynamic NMR Spectroscopy; Academic Press: New York, 1982. [122] Martin, M.L.; Martin, G.J.; Delpuech, J. Practical NMR Spectroscopy; Heyden and Son: London, 1980. [123] Jolly, P.W.; Mynott, R. Adv. Organomet. Chem. 1981, 19, 257. [124] Stengle, T.R.; Langford, C.H. Coord. Chem. Rev. 1967, 2, 349. [125] Forsen, S.; Hoffman, R.A. J. Chem. Phys. 1966, 45, 2059. [126] Cotton, F.A. Acc. Chem. Res. 1968, 1, 257. [127] Faller, J.W. Adv. Organomet. Chem. 1977, 16, 211. [128] Cramer, R. Inorg. Chem. 1965, 4, 445. [129] Kritz, J.; Bouchal, K. J. Organomet. Chem. 1974, ^,255. [130] Miya, S.; Saito, K. Inorg. Chem. 1981, 20, 287. Bibliography 182 131 132 133 134 135 136 137 138 139 140 141 142 Maisonant, A.; Poilblanc, R. Inorg. Chim. Acta. 1978, 29, 203. Herberhold, M.; Kreiter, C.G.; Wiedersatz, G.O. J. Organomet. Chem. 1976, 120, 103. Kang, J.W. Mosely, K. Maitlis, P.M. J. Am. Chem. Soc. 1969, 91, 5970. Moseley, K.; Kang, J.W.; Maitlis, P.M. J. Chem. Soc. (A) 1970, 2875. Moseley, K.; Maitlis, P.M. J. Chem. Soc. (A) 1970, 2884. Holloway, C.E.; Hulley, G.; Johnson, B.F.G.; Lewis, J. J. Chem. Soc. (A) 1969, 53. Holloway, C.E.; Hulley, G.; Johnson, B.F.G.; Lewis, J. J. Chem. Soc. (A) 1970, 1653. Ashley-Smith, J.; Douek, I.; Johnson, B.F.G.; Lewis, J. J. Chem. Soc, Dalton Trans. 1972, 1776. Ashley-Smith, J.; Douek, I.; Johnson, B.F.G.; Lewis, J. J. Chem. Soc, Dalton Trans. 1974, 128. Herberhold, M.; Widersatz, G.O. Chem. Ber. 1976, 109, 3557. Boucher, H.; Bosnich, B. Inorg. Chem. 1977, 16, 717. Bennett, M.A.; Johnson, R.N.; Tomkins, I.B.J. Organomet. Chem. 1977, 133, 231. 143] Albright, T.A.; Hoffman, R.; Thibeault, J.C.; Thorn, D.L. J. Am. Chem. Soc. 1979, 101, 3801. [144] Kaneshima, T.; Kawakami, K.; Tanaka, T. Inorg. Chim. Acta. 1975, 15, 161. Bibliography 183 [145] Kaneshima, T.; Yumoto, Y.; Kawakami, K.; Tanaka, T. Inorg. Chim. Acta. 1976, 18, 29. [146] Kruczynski, L.; LiShingMan, L.K.K.; Takats, J. J. Am. Chem. Soc. 1974, 96, 4006. [147] Wilson, S.T.; Coville, N.J.; Shapely, J.R.; Osborn, J.A. J. Am. Chem. Soc. 1974, 96, 4038. [148] Hoffman, R.; Albright, T.A.; Thorn, D.L. Pure and Appl. Chem. 1978, 50, 1. [149] Albright, T.A. Acc. Chem. Res. 1982, 15, 149. [150] Hogeveen, H.; Nusse, B.J. J. Organomet. Chem. 1979, 171, 237. [151] Chen, M.J.; Feder, H.M. Inorg. Chem. 1979, 18, 1864. [152] Reis, A.H.; Willi, C ; Siegel, S.; Tani, B. Inorg. Chem. 1979, .18 1859. [153] Cocivera, M.; Ferguson, G.; Lalor, F.J.; Szczecinski, P. Organometallics 1982, 1, 1139. [154] a) Tobe, M.L. In Comprehensive Coordination Chemistry; Wilkinson, G.; Gillard, R.D.; McCleverty, J.A., Eds.; Permagon: New York, 1987; Volume 1, Chapter 7, 315; b) Bailar, J.C.Jr. In Comprehensive Coordination Chemistry; Wilkinson, G.; Gillard, R.D.; McCleverty, J.A., Eds.; Permagon: New York, 1987; Volume 1, Chapter 1, 21. [155] a) Eaton, S.S.; Hutchinson, J.R.; Holm, R.H.; Muetterties, E.L. J. Am. Chem. Soc. 1972, 94, 6411; b) Gordon, J.G.; Holm, R.A. J. Am. Chem. Soc. 1970, 92, 5319; c) Allerhand, A.; Gutowsky, H.S.; Jonas, J.; Meinzer, R.A. J. Am. Chem. Soc. 1966, 88, 3185. Bibliography 184 [156] Kleinbaum, D.G.; Kupper, L.L. Applied Regression Analysis and Other Multivari-able Methods; Duxbury: Boston, 1978. [157] Chatt, J.; Venanzi, L.M. J. Chem. Soc. (A) 1957, 4735. [158] Chen, M.J.; Feder, H.M. Inorg. Chem. 1979, 18, No. 7, 1864. [159] Gilman, H.; Wu, T.C. J. Am. Chem. Soc. 1954 , 76, 2502. [160] Still, W.C. J. Org. Chem. 1976, 41, 3063. [161] Corriu, R.J.P.; Guerin, C ; Henner, B.J.L.; Wong Chi Man, W.W.C. Organo-metallics 1988, 7, 237. [162] Corriu, R.J.P.; Perz, R.; Reye, C. Tetrahedron 1983, 39, 999. [163] Corriu, R. J. Organomet. Chem. 1978, 9, 357. Appendix A Bis(/i-t-butylacetato)bis(norbornadiene)dirhodium A number of chiral lanthanide shift reagents are employed for the determination of the optical purity of amines, alcohols, and carboxyhc acids. Some of the shift reagents em-ploy /3-diketonate ligands whose chirality is derived from camphor. In particular the 3-pivaloyl-(-r)-camphorato ligand has been useful in the tris(3-pivaloyl-(-|-)-camphorato-0,0')europium(III) complexes for the determination of the optical purity of primary alcohols [46]. The 3-pivaloyl-(-|-)-camphorato ligand is easily prepared from readily avail-able optically pure camphor [46] and is a /3-diketone making it similar to the employed ligands. In light of these properties, the goal of this part of the project was to prepare the bis(olefin)(3-pivaloyl-(-f )-camphorato-0,0 ')rhodium(I) complex (figure A.57) and to study its abilities as an asymmetric catalyst. The 3-pivaloylcamphor ligand was prepared by refluxing camphor with sodium hy-dride in dimethoxyethane to generate the deprotonated camphor. The next step was the Figure A.57: Target rhodium(I) complex of 3-pivaloyl-(+)-camphor. 185 Appendix A. Bis(p-t-butylacetato)bis(norbomadiene)diThodium 186 addition of pivaloylchloride which produced the 3-pivaloylcamphor ligand and sodium chloride (equation A.28). The reaction mixture was hydrolysed and the ligand purified by vacuum distillation. (A.28) Preparation of the bis(ethene)rhodium(I) complex of 3-pivaloylcamphor (figure A.57, Ri and R 2 = (C2H4)) was attempted by the addition of an aqueous potassium hydrox-ide solution to a mixture of the ligand and the bis(ethene)rhodium(I) chloride dimer at — 78°C. After stirring for 1 h the mixture was allowed to warm to room temperature and was stirred for an additional 1 h. A yellow complex was isolated from the reaction mix-ture, however, it was unstable and decomposed before any analysis could be performed. Repeated attempts gave the same result. In an attempt to prepare a more stable rhodium(I) complex of the ligand the (nor-bornadiene)rhodium(I) chloride dimer was used in place of the bis(ethylene) dimer. By Appendix A. Bis(fi-t-butyla.ceta.to)bis(norbomadiene)dirhodium 187 using the same reaction conditions a stable product was isolated, however, it was not the expected one. The isolated complex was found to be the bis(/z-t-butylacetato)-bis(norbornadiene)dirhodium complex. The crystal structure of the dimer is shown in figure A.58. The elemental analysis, mass spectrometry and X H NMR data (table A.35) for the bis(/x-t-butylacetate)bis(norbornadiene)dirhodium complex were consistent with its for-mulation and structure which was established by x-ray crystallography. The *H NMR spectrum reveals that the complex is fluxional in the NMR time scale, possibly due to a inversion of the eight-membered ring containing the two metal atoms. The NMR spec-trum at room temperature contains a singlet at 3.85 ppm assigned to the olefinic NBD protons, a broad resonance at 3.62 ppm assigned to the protons attached to carbons C(l), C(4), C(8), and C(12), a singlet at 1.01 ppm assigned to the protons attached to the NBD bridgehead protons and a singlet at 1.09 ppm assigned to the methyl groups. At -15°C the olefinic proton resonance at 3.85 ppm splits into two singlets at 3.92 and 3.85 ppm and the resonance at 3.62 ppm splits into two singlets at 3.72 and 3.49 ppm. Assignment of the high field and low field resonances in the low temperature NMR spec-trum were made after inspection of the crystal structure (figure A.58). The protons which are facing the 'inside' of the molecule i.e. olefinic protons H(3), H(5), H(9), and H(13), and aliphatic protons H(4) and H(8) are closer to the carbon-carbon double bond of the opposite NBD ligand. This orientation may deshield the protons on the 'inside' of the molecule causing them to appear further downfield in the NMR spectrum. The mass spectrum of the rhodium dimer contains a peak corresponding to the parent molecular ion at 592 m/z and fragmentation peaks corresponding to the presence of the monomer (296 m/z) and NBDRh (195 m/z). The base peak is due to the presence of the CH(CH 3) 3 fragment ion (57 m/z). Satisfactory elemental analyses data were obtained (carbon expected: 48.66, found: 48.24; hydrogen expected: 5.78, found: 5.79). Appendix A. Bis(p-t-butylacetato)bis(norbornadiene)dirhodium 188 Table A.35: Mass Spectrometry and 1R NMR Data for Bis(//-t-butylacetato)bis(norbornadiene)dirhodium m/z I % 2 6 (ppm) Assignment 3 Protons 4 592 0.2 3.85 a 8 296 13.0 3.62 b 4 195 64.6 1.09 c 18 169 33.3 1.01 d 4 57* 100.0 1The NMR data was obtained at room temperature using toluene-d$ as solvent. ^Intensity (I) is calculated relative to the base peak (*). 3 The assignments are represented as follows: 'a' = olefinic protons, 'b' = the inner and outer methine protons, 'c' = the protons of the t-butyl groups, 'd' = the protons of the bridgehead carbons. 4The number of protons is calculated by relative integral intensity. Appendix A. Bis(p-t-butyla.ceta,to)bis(norbornadiene)diThodium 189 The similar bis(/i-acetato)bis(cyclooctadiene)dirhodium [157] and bis(/z-acetato)bis(nor-bornadiene)dirhodium [158] complexes have previously been reported and are prepared by refluxing a solution of potassium acetate and the appropriate rhodium dimer in acetone. In our reaction it appears as though the complex must have formed and then rearranged to give the dirhodium product. This hypothesis is reasonable as 3-pivaloyl-(-f)-camphor is stable under the conditions used during the reaction. Figure A.58: Crystal structure of bis(/i-t-butj'lacetato)bis(norbornadiene)dirhodium. Appendix B Diphenylboron Complexes of Ligands 9 and 10 Diphenylboron complexes of the 2'-hydroxyacetophenone and 3-benzoyl-(-f )-camphor lig-ands were prepared for structural comparison with the bis(ethene)rhodium(I) complexes of the same ligands. A crystal structure of the (l,5-cyclooctadiene)rhodium(I) com-plex of the 2'-hydroxyacetophenone ligand was performed, however, suitable crystals of a rhodium(I) complex of 3-benzoyl-(-|-)-camphor were not obtained. B . l (2'-acetylphenoxy-0,0')diphenylboron, 21 The (2'-acetylphenoxy-0,0 ')diphenylboron complex was prepared from the reaction of diphenylborinic acid and an excess of 2'-hydroxyacetophenone in diethyl ether (equa-tion B.29). After stirring for 2 h the reaction mixture was filtered, hexane was added and the product was cooled to 5°C. The product crystallized to give yellow platelets which were found to contain one molecule of the 2'-hydroxyacetophenone ligand per two molecules of diphenylboron complex. (B.29) 191 Appendix B. Diphenylboron Complexes of Ligands 9 and 10 192 B.2 (3-Benzoyl-(+)-camphorato-0,0 ^diphenylboron, 22 The diphenylboron complex of 3-benzoyl-(-f-)-camphor was prepared by the reaction of the sodium salt of 3-benzoyl-(+)-camphor and diphenylborinic acid in ether. After stir-ring the mixture at room temperature for 2 h the mixture was filtered, the volume was reduced by solvent evaporation and resultant solution cooled to 5°C. The product crys-tallized as pure yeUowr platelets which were identified as the expected product (discussed below). The crystal structure of the 3-benzoylcamphor diphenylboron complex reveal the presence of two conformation ally different molecules of the same optical configuration (figure B.59 and figure B.60). The two different conformations are illustrated by the stereochemistry about the heterocycle ring formed by chelation. In conformation 'A' the chelate ring is folded so that the bridgehead methyl groups are close to the boron atom and the phenyl groups attached to it. In conformation 'B' the chelate ring is folded in the opposite direction so that the bridgehead methyl groups are further from the boron atom and the phenyl groups attached to it. The elemental analysis of 21 was consistent with the formulation of one molecule of free ligand per two molecules of complex after recrystallization from hexane or ethanol (table B.36). The mass spectrometry and *H NMR data (table B.37) were consistent with the expected structure. A set of multiplets from 6.8 to 7.9 ppm is assigned to the protons of the benzene ring of the ligand and phenyl rings attached to the boron. A singlet at 1.88 ppm is assigned to the methyl group of the ligand. The mass spectrum of 21 contained a peak corresponding to the parent molecular ion (300 m/z) as well as fragment peaks corresponding to the loss of one phenyl group (223 m/z) and two phenyl groups (146 m/z). Appendix B. Diphenylboron Complexes of Ligands 9 and 10 193 The elemental analysis (table B.36), mass spectrometry and a H NMR data (ta-ble B.37) are consistent with the expected structure and formulation for complex 22. A set of resonances from 7.0 to 8.0 ppm with an integral indicating fifteen protons were assigned to the aromatic phenyl protons. A multiplet at 2.6 ppm with an integral indica-tive of one proton is assigned to the proton attached to C(4). Multiplets with integrals indicative of two protons at 1.55 ppm and 1.32 ppm was assigned to the protons of C(6); a multiplet at 1.10 ppm was assigned to the protons of C(5). A singlet at 0.93 ppm with an integral indicative of three protons was assigned to the methyl group of C(2); a singlet with an integral indicative of six protons was assigned to the protons in the methyl groups of C(3) and C(10). The mass spectrum of 22 contained a peak corresponding to the parent molecular ion (420 m/z) as well as fragment peaks corresponding to the loss of a phenyl group (343 m/z). Peaks were also present corresponding to benzoylcamphor (256 m/z) and diphenylboron ions (165 m/z). Figure B.59: Stereoview and crystal structure of (3-beazoyl-(+)-camphorato-0,0 ')-diphenylboron conformation 'A'. Figure B.60: Stereoview and crystal structure of (3-benzoyl-(+)-camphorato-0,0 ')-diphenylboron conformation 'B'. Appendix B. Diphenylboron Complexes of Ligands 9 and 10 Table B.36: Elemental Analysis Data for The Boron Complexes Compound Carbon Hydrogen 21 -1/2 Ligand Expected 78.28 5.75 Found 1 78.45 5.66 Found 2 78.04 5.82 22 Expected 82.86 6.96 Found 82.60 6.94 Recrystallized from hexane. 2Recrystallized from ethanol. Appendix B. Diphenylboron Complexes of Ligands 9 and 10 197 Table B.37: : H NMR Data for the Boron Complexes 1 £(ppm) Complex 21 Assignment Protons 2 £(ppm) Complex 22 Assignment 3 Protons 6.75 -7.9 aromatic 14 7.0 - 8.0 aromatic 15 1.88 methyl 3 2.6 C(4) 1 1.55, 1.32 C(6) 2 1.10 C(5) 2 0.93 C(9) 3 0.43 C(8), C(10) 6 aThe NMR data was obtained at room temperature using CDCI3 as solvent. 2 The number of protons is calculated by relative integral intensity. 3The assignment lists the carbon to which the appropriate protons are attached (fig-ure B.59). Appendix C F A C O O H and Related Esters The success in the use of metal complexes of ligands prepared from a-N,N-dimethylethyl-arninoferrocene (FA) as catalysts for asymmetric hydrogenation prompted attempts to prepare a FA-/3-diketonate ligand. Unfortunately, the preparation of the target ligand was not achieved, however, the attempts gave rise to some interesting compounds. The route chosen to prepare the target ligands discussed in Chapter 4 proceeded via the preparation of FACOOH. This compound was found to have some interesting properties. The *H NMR spectrum of FACOOH indicates that the molecule is fluxional in the NMR time scale, and further investigation revealed the temperature dependence of the motion (figure C.61). The observed effect appears to be due to hindrance of the inversion at the nitrogen center. The cause of the hindrance is attributed to a hydrogen bond between the carboxylic hydrogen and the nitrogen. The methyl ester of FACOOH, FACOOMe, does not show this effect and both N-methyl groups appear as equivalent on the NMR time scale. An interesting effect which is present at low temperatures, when the nitrogen inversion is slow, is coupling between the hydroxyl proton and the protons of the methyl group on the nitrogen. This coupling must be propagated through the nitrogen-hydrogen bond, and its presence gives an indication of the strength of this interaction. 198 Appendix C: FACOOH and Related Esters 199 Figure C.61: Variable temperature : H NMR of the N-methyl region of FACOOH. Appendix C. FACOOH and Related Esters 200 Preparation of esters from FACOOH was difficult and esterification attempts made by refluxing the acid with ethanol or methanol showed no reaction after several hours. Attempts to esterify FACOOH were then made using the method reported by Alvarez et al for the esterification of sterically hindered carboxylic acids. However use of this technique did not give the corresponding ester of FACOOH but instead produced an ester in which the dimethylamine group had been replaced by a hydroxyl group (equa-tion C.30). Replacement of the dimethylamine group of FA complexes with a hydroxyl group has been accomplished by first replacing it with an acetate group, by reaction with acetic anhydride, followed by base hydrolysis [113]. Attempts to prepare the methyl ester yielded 2-(a-hydroxyethylferrocene)carboxylic acid methyl ester (FCMe) and attempts to prepare the ethyl ester yielded 2-(a-hydroxylethylferrocene)carboxylic acid ethyl ester (FCEt). (C.30) NMe2 COOH RI, K 2 C 0 3 DMA, H 20 Appendix C. FACOOH and Related Esters 201 The 1 H NMR spectrum of FCMe indicated that the hydroxyl proton is strongly coupled to proton 'a' (figure C.62). The resonance due to proton 'a' appears as a doublet of quartets at 5.00 ppm and the hydroxyl proton appears as a well resolved doublet at 4.36 ppm (figure C.62). To confirm this assignment the sample was shaken with D 2 0 to give the hydroxyl proton 'decoupled' spectrum; now the 'a' proton resonance is a quartet and a peak appears at 4.5 ppm due to the HDO formed from the exchange reaction. The strong coupling interaction between the hydroxyl proton and proton 'a' is an indication of a possible hydrogen bond between the hydroxyl proton and the ketone oxygen. This interaction would decrease the exchange between hydroxyl protons by anchoring them in a seven-membered ring anlogous to the seven-membered ring evident in FACOOH. A crystal structure of FCMe was obtained and is shown in figure C.63. The crystal structure does not show an intramolecular interaction between the hydroxyl proton, however, an intermolecular hydrogen bond is present between the hydroxyl proton and the methoxy oxygen; it is not possible to rule out the presence of the hydroxyl-ketone interaction in solution even though the solid state prefers the hydroxyl-methoxide oxygen interaction. The elemental analysis (table C.38) mass spectrometry and *H NMR data (table C.39) obtained for FCMe are consistent with the expected structure and formulation. A doublet of quartets with an integral indicative of one proton is assigned to proton 'a' (J = 6.7 Hz); 'a' is coupled to the neighbouring methyl group and the hydroxyl proton. A multiplet at 4.70 ppm with an integral indicative of one proton is assigned to the ferrocenyl proton 'b' which is deshielded with respect to the other ferrocenyl protons because of its proximity to the carbonyl group. Multiplets at 4.40 and 4.27 ppm are assigned to ferrocenyl protons 'c' and 'e'. A doublet at 4.36 ppm with an integral indicative of one proton is assigned to the hydroxyl proton; this assignment is verified by exchange with D 2 0 and will be discussed in more detail later. A singlet at 4.12 ppm with an integral indicative of five Appendix C. FACOOH and Related Esters 202 protons is assigned to the unsubstituted ferrocenyl ring protons T . A singlet at 3.78 ppm with an integral indicative of three protons is assigned to the ester methyl group protons 'g'. A doublet at 1.45 ppm with an integral indicative of three protons is assigned to the 'h' methyl protons which are coupled to proton 'a' (J = 6.3 Hz). The mass spectrum of FCMe contains a peak corresponding to the parent molecular ion (288 m/z) as well as fragment peaks corresponding to the loss of H 2 O (270 m/z), C H 3 O H (256 m/z), Cp (223 m/z) and FeCp (119 m/z). The ethyl ester, FCEt, has a similar X H NMR spectrum to FCMe (table C.40). The differences are due to the ethyl group which gives rise to a triplet at 1.36 ppm due to the methyl 'h' protons and a quartet at 4.43 ppm due to the methylene 'e' protons. The quartet is obscurred by the ferrocenyl protons which also resonate in this region of the spectrum. The mass spectrum of FCEt contains a peak corresponding to the parent molecular ion (302 m/z) as well as fragment peaks corresponding to the loss of H 2 O (234 m/z), H O C H C H 3 or HOCH 2 CH 3 (256 m/z), Cp (237 m/z) and FeCp (119 m/z). Appendix C. FACOOH and Related Esters 204 Figure C.63: Crystal structure of 2-(a-hydroxyethylferrocene)carboxylic acid methyl es-ter. Appendix C. FACOOH and Related Esters Table C.38: Elemental Analyses Data for FCMe and FCEt Complex Carbon Hydrogen Oxygen— FCMe Expected 58.36 5.60 16.66 Found 58.07 5.86 16.69 FCEt Expected 59.63 6.00 15.88 Found 59.62 5.95 16.00 Appendix C. FACOOH and Related Esters 206 Table C.39: Mass Spectrometry and 1 H NMR Data for FCMe 1 m/z I (%) 2 6(ppm) Assignment Protons 3 288 49.96 5.00(dbl. of quartets,6.7 Hz) a 270 25.50 4.70 (multiplet) b i 256 22.67 4.40 (multiplet) c i 223 15.93 4.36 (doublet.6.7 Hz) d 212 32.72 4.27 (multiplet) e i 152 15.80 4.12 f 5 138 16.27 3.78 g 3 119* 100.0 1.45 (doublet, 6.7 Hz) h 3 JThe NMR data was obtained at room temperature using CDCI3 as solvent. 2Intensity (I) is calculated relative to the base peak (*). 3The number of protons is calculated by relative integral intensity. CH 3 Appendix C. FACOOH and Related Esters Table C.40: Mass Spectrometry and lE NMR Data for FCEt 1 m/z I (%) 2 c^ (ppm) Assignment Protons 3 302 48.67 4.97(dbl. of quartets,7.2 Hz) a 1 284 38.82 4.76(multiplet) b ' 1 256 39.95 4.52(doublet,7.2 Hz) c 1 237 20.15 4.43 d,e 4 219 14.25 4.17 f 5 212 18.80 1.48(doublet,6.3 Hz) g 3 119* 100.0 1.36(triplet,6.3 Hz) h • 3 1The NMR data was obtained at room temperature using CDCI3 as solvent. intensity (I) is calculated realative to the base peak (*). 3The number of protons is calculated by relative integral intensity. 9 Appendix D Base-Catalysed Hydrosilylation As mentioned in Chapter 4, the hydrosilylation reaction is useful for the preparation of oganosilicon reagents and for the reduction of carbon-carbon and carbon-oxygen multiple bonds [101, 102]. The ability of chiral catalysts to affect the asymmetric hydrosilylation of unsaturated molecules has increased the amount of interest in this area [24]. Usually the catalysts used for the hydrosilylation reaction are based on expensive metals, such as rhodium, platinum and palladium [24, 104]. This chapter deals with a new catalytic hydrosilylation reaction, which effects high conversion of silane and ketone to silyl ether, without the use of a platinum metal catalyst. Bases such as sodium hydride, sodium methoxide or lithium diisopropylamide were found to catalyse the hydrosilylation of ketones by diphenylsilane. The ketone hydrosi-lylation reaction affords high yields of the corresponding silyl ether. The reaction also shows some selectivity in the hydrosilylation of 2-cyclohexen-l-one. The hydrosilylation reaction is performed by adding the base to a mixture of diphenyl-silane and ketone with or without solvent. The use of dilute conditions slows the rate but does not affect the overall yield. The reaction does not seem to be affected by aromatic substituents1 but is affected by steric hindrance in the substrate (table D.41). The base-catalysed hydrosilylation of 2-cyclohexen-l-one by diphenylsilane shows preference for either mode of reaction (1,2 addition or 1,4 addition) depending on the 1 Wilkinson's catalyst has been found to be an effective catalyst for the hydrosilylation of aliphatic ketones but the reaction is slowed considerably by the presence of an aromatic substituent on the ketone. It is then necessary to heat the reaction mixture to 60°C to effect reaction. 208 Appendix D. Base-Catalysed Hydrosilylation 209 initial molar ratio of silane to enone (table D.42). High enone to silane ratios favour 1,4-addition to give product A whereas low ratios favour 1,2 addition to give product B after hydrolysis. Compound C, which is the product of the addition of diphenylsilane across both double bonds, is present in greater amounts when high ratios of silane : enone are used and is less abundant when using dilute reaction conditions. An attempt was made to effect the asymmetric hydrosilylation of acetophenone using a chiral base. The (#)-(•+)- a-methylbenzylamine was chosen as the source of chirality and lithium methylbenzylamide was prepared from it by reacting the amine with butyllithium. The lithium methylbenzylamide was prepared by using the same method employed for the preparation of lithium diisopropylamide using THF as solvent. The catalytic reaction was carried out by adding a mixture of ketone and silane to a solution of the chiral amide. Use of the optically active base did give rise to asymmetry in the hydrosilylation product, however, optical yields are low. For a ratio of amide : ketone : silane of 1 : 100 : 105 a chemical yield of 91.3% and an optical yield of (-)0.07% (e.e.) resulted after 15 h. Doubling the amount of amide present gave a chemical yield of 96.4% and an optical yield of (-)0.50% (e.e.) after 40 h. Optical yields were calculated for the hydrolysis product, i.e. phenethyl alcohol, by optical rotation measurements. Although the optical yields are low, this reaction may have potential in asymmetric catalysis providing a better combination of base and ketone are found; this result does demonstrate such a possibility. D . l Proposed Mechanism Alkali metal triaikylsilanes have been reported to add across carbon-oxygen double bonds to give the corresponding silyl ether [159]. The addition of lithium trimethylsilane to enones also has been reported to proceed selectively in a 1,4 mode of addition [160]. These reactions do not account for the present results since they are stoichiometric addition Appendix D. Base-Catalysed Hydrosilylation 210 Table D.41: Base-Catalysed Hydrosilylation Results Ketone Conditions 1 Alcohol Conversion(%) 2 0 II Or 20° C, 20 h O H ^ C H C H j Or 99 0 II or 20c C, 24 h O H 1 _ C H C H 2 C H 2 C H j Or 95 0 20° C, 24 h O H ^ C H C H ( C H 3 ) 2 Or 79 0 C M J ? C H 2 C M ( C H J ) J 20° C, 24 h O H 1 C H 3 C H C H 2 C H ( C H j ) j 97 1The reactions are carried out in the absence of solvent starting with a, 1.0 : 1.1 ratio of ketone to silane to which the base is added (1 mol percent). The base employed here is sodium hydride, the other bases afford comparable results. The reactions were highly exothermic. 2Determined on the basis of gas chromatographic data. Appendix D. Base-Catalysed Hydrosilylation Table D.42: Base-Catalysed Hydrosilylation of 2-Cyclohexen-1-one 0 0 OH OH 1) NaH (catalyst) X JL JL 0 + p h ^ 2 ) K 2 c o , ( , q ) » 0+ 0 + 0 A B C Enone:Silane Conversion (%) 1 A B C Conditions 2 2:1 36.6 47.4 38.4 14.2 100 h, neat 1:1 81.6 11.4 52.5 36.1 72 h, neat 1:2 100 - 65.4 34.6 100 h, neat 2:3 90.0 9.4 74.6 16.0 75 h, THF 1:2 94.8 7.9 78.1 14.0 100 h, THF 1 Determined on the basis of gas chromatographic data. 2Reactions cooled to 0°C during addition of sodium hydride (base : ketone of 1 all reactions were highly exothermic. Appendix D. Base-Catalysed Hydrosilylation 212 reactions. An explanation for the base-catalysed reaction may be found in the comparison of the system with the known organosihcon-flouride activation reaction. Organosilicon com-pounds are known to form anionic species with enhanced reactivity towards nucleophiles in the presence of flouride ion [161]. When the organosilicon compound is a silane, fluo-ride activation results in an active hydrogen on the silicon atom; this effect is explained by a weakening of the silicon-hydrogen bond due to coordination of the flouride ion to the silicon atom [162]. A five coordinate anionic silane species formed by the addition of cesium flouride to diphenylsilane has been reported to add across carbon-oxygen double bonds (figure D.64) [163]. Although this reaction is stoichiometric, it is possible that the presence of a five-coordinate silane anion may be responsible for the reaction discussed here, however in this system an H " ion must be activating the silicon species. A mech-anism in which a five coordinate anionic silane species is formed and regenerated after addition of the silane to the ketone could then explain the catalytic reaction. A possible mechanism for this reaction is outlined in figure D.65. The initial step of the mechanism involves the production of a five coordinate silyl anion intermediate (A) by the reaction between sodium hydride and diphenylsilane. The reaction could then proceed by oxygen coordination to the silane (B). In order for the reaction to proceed B must react to regenerate the catalyst. The catalyst regenerations may occur from B or from the silyl ether anion derivative of B formed by insertion of a hydride into the ketone bond. The most likely route for catalyst regeneration is the exchange of a hydride from B, or its silyl ether, with a molecule of diphenylsilane to give A, and the isolated silyl ether product, C. eudix D. Base-Catalysed Hydrosilylation 213 OEt EtO | E t 0 - r H C s + 0, ' C — R \ R OEt I H E t O - S i - 0 - C R 2 I OEt Figure D.64: Addition of a five-coordinate silyl anion to a ketone. Ph 0 - C - C H 3 I Ph (O Ph H Si P h ' X H 0""~CH3 Ph 2 SiH 2 Figure D.65: Possible mechanism for base-catalysed hydrosilylation. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0059733/manifest

Comment

Related Items